Merge branch 'main' into litellm_dev_04_22_2025_p1

docs/my-website/docs/observability/agentops_integration.md

@@ -0,0 +1,83 @@
+# 🖇️ AgentOps - LLM Observability Platform
+
+:::tip
+
+This is community maintained. Please make an issue if you run into a bug:
+https://github.com/BerriAI/litellm
+
+:::
+
+[AgentOps](https://docs.agentops.ai) is an observability platform that enables tracing and monitoring of LLM calls, providing detailed insights into your AI operations.
+
+## Using AgentOps with LiteLLM
+
+LiteLLM provides `success_callbacks` and `failure_callbacks`, allowing you to easily integrate AgentOps for comprehensive tracing and monitoring of your LLM operations.
+
+### Integration
+
+Use just a few lines of code to instantly trace your responses **across all providers** with AgentOps:
+Get your AgentOps API Keys from https://app.agentops.ai/
+```python
+import litellm
+
+# Configure LiteLLM to use AgentOps
+litellm.success_callback = ["agentops"]
+
+# Make your LLM calls as usual
+response = litellm.completion(
+    model="gpt-3.5-turbo",
+    messages=[{"role": "user", "content": "Hello, how are you?"}],
+)
+```
+
+Complete Code:
+
+```python
+import os
+from litellm import completion
+
+# Set env variables
+os.environ["OPENAI_API_KEY"] = "your-openai-key"
+os.environ["AGENTOPS_API_KEY"] = "your-agentops-api-key"
+
+# Configure LiteLLM to use AgentOps
+litellm.success_callback = ["agentops"]
+
+# OpenAI call
+response = completion(
+    model="gpt-4",
+    messages=[{"role": "user", "content": "Hi 👋 - I'm OpenAI"}],
+)
+
+print(response)
+```
+
+### Configuration Options
+
+The AgentOps integration can be configured through environment variables:
+
+- `AGENTOPS_API_KEY` (str, optional): Your AgentOps API key
+- `AGENTOPS_ENVIRONMENT` (str, optional): Deployment environment (defaults to "production")
+- `AGENTOPS_SERVICE_NAME` (str, optional): Service name for tracing (defaults to "agentops")
+
+### Advanced Usage
+
+You can configure additional settings through environment variables:
+
+```python
+import os
+
+# Configure AgentOps settings
+os.environ["AGENTOPS_API_KEY"] = "your-agentops-api-key"
+os.environ["AGENTOPS_ENVIRONMENT"] = "staging"
+os.environ["AGENTOPS_SERVICE_NAME"] = "my-service"
+
+# Enable AgentOps tracing
+litellm.success_callback = ["agentops"]
+```
+
+### Support
+
+For issues or questions, please refer to:
+- [AgentOps Documentation](https://docs.agentops.ai)
+- [LiteLLM Documentation](https://docs.litellm.ai) 

docs/my-website/docs/proxy/config_settings.md

@@ -299,6 +299,9 @@ router_settings:
 |------|-------------|
 | ACTIONS_ID_TOKEN_REQUEST_TOKEN | Token for requesting ID in GitHub Actions
 | ACTIONS_ID_TOKEN_REQUEST_URL | URL for requesting ID token in GitHub Actions
+| AGENTOPS_ENVIRONMENT | Environment for AgentOps logging integration
+| AGENTOPS_API_KEY | API Key for AgentOps logging integration
+| AGENTOPS_SERVICE_NAME | Service Name for AgentOps logging integration
 | AISPEND_ACCOUNT_ID | Account ID for AI Spend
 | AISPEND_API_KEY | API Key for AI Spend
 | ALLOWED_EMAIL_DOMAINS | List of email domains allowed for access

docs/my-website/sidebars.js

@@ -411,6 +411,7 @@ const sidebars = {
       type: "category",
       label: "Logging & Observability",
       items: [
+        "observability/agentops_integration",
         "observability/langfuse_integration",
         "observability/lunary_integration",
         "observability/mlflow",

litellm/__init__.py

@@ -113,6 +113,7 @@ _custom_logger_compatible_callbacks_literal = Literal[
     "pagerduty",
     "humanloop",
     "gcs_pubsub",
+    "agentops",
     "anthropic_cache_control_hook",
 ]
 logged_real_time_event_types: Optional[Union[List[str], Literal["*"]]] = None

litellm/integrations/_types/open_inference.py

@@ -45,6 +45,14 @@ class SpanAttributes:
     """
     The name of the model being used.
     """
+    LLM_PROVIDER = "llm.provider"
+    """
+    The provider of the model, such as OpenAI, Azure, Google, etc.
+    """
+    LLM_SYSTEM = "llm.system"
+    """
+    The AI product as identified by the client or server
+    """
     LLM_PROMPTS = "llm.prompts"
     """
     Prompts provided to a completions API.
@@ -65,15 +73,40 @@ class SpanAttributes:
     """
     Number of tokens in the prompt.
     """
+    LLM_TOKEN_COUNT_PROMPT_DETAILS_CACHE_WRITE = "llm.token_count.prompt_details.cache_write"
+    """
+    Number of tokens in the prompt that were written to cache.
+    """
+    LLM_TOKEN_COUNT_PROMPT_DETAILS_CACHE_READ = "llm.token_count.prompt_details.cache_read"
+    """
+    Number of tokens in the prompt that were read from cache.
+    """
+    LLM_TOKEN_COUNT_PROMPT_DETAILS_AUDIO = "llm.token_count.prompt_details.audio"
+    """
+    The number of audio input tokens presented in the prompt
+    """
     LLM_TOKEN_COUNT_COMPLETION = "llm.token_count.completion"
     """
     Number of tokens in the completion.
     """
+    LLM_TOKEN_COUNT_COMPLETION_DETAILS_REASONING = "llm.token_count.completion_details.reasoning"
+    """
+    Number of tokens used for reasoning steps in the completion.
+    """
+    LLM_TOKEN_COUNT_COMPLETION_DETAILS_AUDIO = "llm.token_count.completion_details.audio"
+    """
+    The number of audio input tokens generated by the model
+    """
     LLM_TOKEN_COUNT_TOTAL = "llm.token_count.total"
     """
     Total number of tokens, including both prompt and completion.
     """
 
+    LLM_TOOLS = "llm.tools"
+    """
+    List of tools that are advertised to the LLM to be able to call
+    """
+
     TOOL_NAME = "tool.name"
     """
     Name of the tool being used.
@@ -112,6 +145,19 @@ class SpanAttributes:
     The id of the user
     """
 
+    PROMPT_VENDOR = "prompt.vendor"
+    """
+    The vendor or origin of the prompt, e.g. a prompt library, a specialized service, etc.
+    """
+    PROMPT_ID = "prompt.id"
+    """
+    A vendor-specific id used to locate the prompt.
+    """
+    PROMPT_URL = "prompt.url"
+    """
+    A vendor-specific url used to locate the prompt.
+    """
+
 
 class MessageAttributes:
     """
@@ -151,6 +197,10 @@ class MessageAttributes:
     The JSON string representing the arguments passed to the function
     during a function call.
     """
+    MESSAGE_TOOL_CALL_ID = "message.tool_call_id"
+    """
+    The id of the tool call.
+    """
 
 
 class MessageContentAttributes:
@@ -186,6 +236,25 @@ class ImageAttributes:
     """
 
 
+class AudioAttributes:
+    """
+    Attributes for audio
+    """
+
+    AUDIO_URL = "audio.url"
+    """
+    The url to an audio file
+    """
+    AUDIO_MIME_TYPE = "audio.mime_type"
+    """
+    The mime type of the audio file
+    """
+    AUDIO_TRANSCRIPT = "audio.transcript"
+    """
+    The transcript of the audio file
+    """
+
+
 class DocumentAttributes:
     """
     Attributes for a document.
@@ -257,6 +326,10 @@ class ToolCallAttributes:
     Attributes for a tool call
     """
 
+    TOOL_CALL_ID = "tool_call.id"
+    """
+    The id of the tool call.
+    """
     TOOL_CALL_FUNCTION_NAME = "tool_call.function.name"
     """
     The name of function that is being called during a tool call.
@@ -268,6 +341,18 @@ class ToolCallAttributes:
     """
 
 
+class ToolAttributes:
+    """
+    Attributes for a tools
+    """
+
+    TOOL_JSON_SCHEMA = "tool.json_schema"
+    """
+    The json schema of a tool input, It is RECOMMENDED that this be in the
+    OpenAI tool calling format: https://platform.openai.com/docs/assistants/tools
+    """
+
+
 class OpenInferenceSpanKindValues(Enum):
     TOOL = "TOOL"
     CHAIN = "CHAIN"
@@ -284,3 +369,21 @@ class OpenInferenceSpanKindValues(Enum):
 class OpenInferenceMimeTypeValues(Enum):
     TEXT = "text/plain"
     JSON = "application/json"
+
+
+class OpenInferenceLLMSystemValues(Enum):
+    OPENAI = "openai"
+    ANTHROPIC = "anthropic"
+    COHERE = "cohere"
+    MISTRALAI = "mistralai"
+    VERTEXAI = "vertexai"
+
+
+class OpenInferenceLLMProviderValues(Enum):
+    OPENAI = "openai"
+    ANTHROPIC = "anthropic"
+    COHERE = "cohere"
+    MISTRALAI = "mistralai"
+    GOOGLE = "google"
+    AZURE = "azure"
+    AWS = "aws"

litellm/integrations/agentops/__init__.py

@@ -0,0 +1,3 @@
+from .agentops import AgentOps
+
+__all__ = ["AgentOps"] 

litellm/integrations/agentops/agentops.py

@@ -0,0 +1,118 @@
+"""
+AgentOps integration for LiteLLM - Provides OpenTelemetry tracing for LLM calls
+"""
+import os
+from dataclasses import dataclass
+from typing import Optional, Dict, Any
+from litellm.integrations.opentelemetry import OpenTelemetry, OpenTelemetryConfig
+from litellm.llms.custom_httpx.http_handler import _get_httpx_client
+
+@dataclass
+class AgentOpsConfig:
+    endpoint: str = "https://otlp.agentops.cloud/v1/traces"
+    api_key: Optional[str] = None
+    service_name: Optional[str] = None
+    deployment_environment: Optional[str] = None
+    auth_endpoint: str = "https://api.agentops.ai/v3/auth/token"
+
+    @classmethod
+    def from_env(cls):
+        return cls(
+            endpoint="https://otlp.agentops.cloud/v1/traces",
+            api_key=os.getenv("AGENTOPS_API_KEY"),
+            service_name=os.getenv("AGENTOPS_SERVICE_NAME", "agentops"),
+            deployment_environment=os.getenv("AGENTOPS_ENVIRONMENT", "production"),
+            auth_endpoint="https://api.agentops.ai/v3/auth/token"
+        )
+
+class AgentOps(OpenTelemetry):
+    """
+    AgentOps integration - built on top of OpenTelemetry
+
+    Example usage:
+        ```python
+        import litellm
+        
+        litellm.success_callback = ["agentops"]
+
+        response = litellm.completion(
+            model="gpt-3.5-turbo",
+            messages=[{"role": "user", "content": "Hello, how are you?"}],
+        )
+        ```
+    """
+    def __init__(
+        self,
+        config: Optional[AgentOpsConfig] = None,
+    ):
+        if config is None:
+            config = AgentOpsConfig.from_env()
+
+        # Prefetch JWT token for authentication
+        jwt_token = None
+        project_id = None
+        if config.api_key:
+            try:
+                response = self._fetch_auth_token(config.api_key, config.auth_endpoint)
+                jwt_token = response.get("token")
+                project_id = response.get("project_id")
+            except Exception:
+                pass
+
+        headers = f"Authorization=Bearer {jwt_token}" if jwt_token else None
+        
+        otel_config = OpenTelemetryConfig(
+            exporter="otlp_http",
+            endpoint=config.endpoint,
+            headers=headers
+        )
+
+        # Initialize OpenTelemetry with our config
+        super().__init__(
+            config=otel_config,
+            callback_name="agentops"
+        )
+
+        # Set AgentOps-specific resource attributes
+        resource_attrs = {
+            "service.name": config.service_name or "litellm",
+            "deployment.environment": config.deployment_environment or "production",
+            "telemetry.sdk.name": "agentops",
+        }
+        
+        if project_id:
+            resource_attrs["project.id"] = project_id
+            
+        self.resource_attributes = resource_attrs
+
+    def _fetch_auth_token(self, api_key: str, auth_endpoint: str) -> Dict[str, Any]:
+        """
+        Fetch JWT authentication token from AgentOps API
+        
+        Args:
+            api_key: AgentOps API key
+            auth_endpoint: Authentication endpoint
+            
+        Returns:
+            Dict containing JWT token and project ID
+        """
+        headers = {
+            "Content-Type": "application/json",
+            "Connection": "keep-alive",
+        }
+        
+        client = _get_httpx_client()
+        try:
+            response = client.post(
+                url=auth_endpoint,
+                headers=headers,
+                json={"api_key": api_key},
+                timeout=10
+            )
+            
+            if response.status_code != 200:
+                raise Exception(f"Failed to fetch auth token: {response.text}")
+            
+            return response.json()
+        finally:
+            client.close() 

litellm/integrations/arize/_utils.py

@@ -1,3 +1,4 @@
+import json
 from typing import TYPE_CHECKING, Any, Optional, Union
 
 from litellm._logging import verbose_logger
@@ -12,36 +13,141 @@ else:
     Span = Any
 
 
-def set_attributes(span: Span, kwargs, response_obj):
+def cast_as_primitive_value_type(value) -> Union[str, bool, int, float]:
+    """
+    Converts a value to an OTEL-supported primitive for Arize/Phoenix observability.
+    """
+    if value is None:
+        return ""
+    if isinstance(value, (str, bool, int, float)):
+        return value
+    try:
+        return str(value)
+    except Exception:
+        return ""
+
+
+def safe_set_attribute(span: Span, key: str, value: Any):
+    """
+    Sets a span attribute safely with OTEL-compliant primitive typing for Arize/Phoenix.
+    """
+    primitive_value = cast_as_primitive_value_type(value)
+    span.set_attribute(key, primitive_value)
+
+
+def set_attributes(span: Span, kwargs, response_obj):  # noqa: PLR0915
+    """
+    Populates span with OpenInference-compliant LLM attributes for Arize and Phoenix tracing.
+    """
     from litellm.integrations._types.open_inference import (
         MessageAttributes,
         OpenInferenceSpanKindValues,
         SpanAttributes,
+        ToolCallAttributes,
     )
 
     try:
+        optional_params = kwargs.get("optional_params", {})
+        litellm_params = kwargs.get("litellm_params", {})
         standard_logging_payload: Optional[StandardLoggingPayload] = kwargs.get(
             "standard_logging_object"
         )
+        if standard_logging_payload is None:
+            raise ValueError("standard_logging_object not found in kwargs")
 
         #############################################
         ############ LLM CALL METADATA ##############
         #############################################
 
-        if standard_logging_payload and (
-            metadata := standard_logging_payload["metadata"]
-        ):
-            span.set_attribute(SpanAttributes.METADATA, safe_dumps(metadata))
+        # Set custom metadata for observability and trace enrichment.
+        metadata = (
+            standard_logging_payload.get("metadata")
+            if standard_logging_payload
+            else None
+        )
+        if metadata is not None:
+            safe_set_attribute(span, SpanAttributes.METADATA, safe_dumps(metadata))
 
         #############################################
         ########## LLM Request Attributes ###########
         #############################################
 
-        # The name of the LLM a request is being made to
+        # The name of the LLM a request is being made to.
         if kwargs.get("model"):
-            span.set_attribute(SpanAttributes.LLM_MODEL_NAME, kwargs.get("model"))
+            safe_set_attribute(
+                span,
+                SpanAttributes.LLM_MODEL_NAME,
+                kwargs.get("model"),
+            )
 
-        span.set_attribute(
+        # The LLM request type.
+        safe_set_attribute(
+            span,
+            "llm.request.type",
+            standard_logging_payload["call_type"],
+        )
+
+        # The Generative AI Provider: Azure, OpenAI, etc.
+        safe_set_attribute(
+            span,
+            SpanAttributes.LLM_PROVIDER,
+            litellm_params.get("custom_llm_provider", "Unknown"),
+        )
+
+        # The maximum number of tokens the LLM generates for a request.
+        if optional_params.get("max_tokens"):
+            safe_set_attribute(
+                span,
+                "llm.request.max_tokens",
+                optional_params.get("max_tokens"),
+            )
+
+        # The temperature setting for the LLM request.
+        if optional_params.get("temperature"):
+            safe_set_attribute(
+                span,
+                "llm.request.temperature",
+                optional_params.get("temperature"),
+            )
+
+        # The top_p sampling setting for the LLM request.
+        if optional_params.get("top_p"):
+            safe_set_attribute(
+                span,
+                "llm.request.top_p",
+                optional_params.get("top_p"),
+            )
+
+        # Indicates whether response is streamed.
+        safe_set_attribute(
+            span,
+            "llm.is_streaming",
+            str(optional_params.get("stream", False)),
+        )
+
+        # Logs the user ID if present.
+        if optional_params.get("user"):
+            safe_set_attribute(
+                span,
+                "llm.user",
+                optional_params.get("user"),
+            )
+
+        # The unique identifier for the completion.
+        if response_obj and response_obj.get("id"):
+            safe_set_attribute(span, "llm.response.id", response_obj.get("id"))
+
+        # The model used to generate the response.
+        if response_obj and response_obj.get("model"):
+            safe_set_attribute(
+                span,
+                "llm.response.model",
+                response_obj.get("model"),
+            )
+
+        # Required by OpenInference to mark span as LLM kind.
+        safe_set_attribute(
+            span,
             SpanAttributes.OPENINFERENCE_SPAN_KIND,
             OpenInferenceSpanKindValues.LLM.value,
         )
@@ -50,77 +156,132 @@ def set_attributes(span: Span, kwargs, response_obj):
         # for /chat/completions
         # https://docs.arize.com/arize/large-language-models/tracing/semantic-conventions
         if messages:
-            span.set_attribute(
+            last_message = messages[-1]
+            safe_set_attribute(
+                span,
                 SpanAttributes.INPUT_VALUE,
-                messages[-1].get("content", ""),  # get the last message for input
+                last_message.get("content", ""),
             )
 
-            # LLM_INPUT_MESSAGES shows up under `input_messages` tab on the span page
+            # LLM_INPUT_MESSAGES shows up under `input_messages` tab on the span page.
             for idx, msg in enumerate(messages):
-                # Set the role per message
-                span.set_attribute(
-                    f"{SpanAttributes.LLM_INPUT_MESSAGES}.{idx}.{MessageAttributes.MESSAGE_ROLE}",
-                    msg["role"],
+                prefix = f"{SpanAttributes.LLM_INPUT_MESSAGES}.{idx}"
+                # Set the role per message.
+                safe_set_attribute(
+                    span, f"{prefix}.{MessageAttributes.MESSAGE_ROLE}", msg.get("role")
                 )
-                # Set the content per message
-                span.set_attribute(
-                    f"{SpanAttributes.LLM_INPUT_MESSAGES}.{idx}.{MessageAttributes.MESSAGE_CONTENT}",
+                # Set the content per message.
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{MessageAttributes.MESSAGE_CONTENT}",
                     msg.get("content", ""),
                 )
 
-        if standard_logging_payload and (
-            model_params := standard_logging_payload["model_parameters"]
-        ):
+        # Capture tools (function definitions) used in the LLM call.
+        tools = optional_params.get("tools")
+        if tools:
+            for idx, tool in enumerate(tools):
+                function = tool.get("function")
+                if not function:
+                    continue
+                prefix = f"{SpanAttributes.LLM_TOOLS}.{idx}"
+                safe_set_attribute(
+                    span, f"{prefix}.{SpanAttributes.TOOL_NAME}", function.get("name")
+                )
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{SpanAttributes.TOOL_DESCRIPTION}",
+                    function.get("description"),
+                )
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{SpanAttributes.TOOL_PARAMETERS}",
+                    json.dumps(function.get("parameters")),
+                )
+
+        # Capture tool calls made during function-calling LLM flows.
+        functions = optional_params.get("functions")
+        if functions:
+            for idx, function in enumerate(functions):
+                prefix = f"{MessageAttributes.MESSAGE_TOOL_CALLS}.{idx}"
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{ToolCallAttributes.TOOL_CALL_FUNCTION_NAME}",
+                    function.get("name"),
+                )
+
+        # Capture invocation parameters and user ID if available.
+        model_params = (
+            standard_logging_payload.get("model_parameters")
+            if standard_logging_payload
+            else None
+        )
+        if model_params:
             # The Generative AI Provider: Azure, OpenAI, etc.
-            span.set_attribute(
-                SpanAttributes.LLM_INVOCATION_PARAMETERS, safe_dumps(model_params)
+            safe_set_attribute(
+                span,
+                SpanAttributes.LLM_INVOCATION_PARAMETERS,
+                safe_dumps(model_params),
             )
 
             if model_params.get("user"):
                 user_id = model_params.get("user")
                 if user_id is not None:
-                    span.set_attribute(SpanAttributes.USER_ID, user_id)
+                    safe_set_attribute(span, SpanAttributes.USER_ID, user_id)
 
         #############################################
         ########## LLM Response Attributes ##########
-        # https://docs.arize.com/arize/large-language-models/tracing/semantic-conventions
         #############################################
-        if hasattr(response_obj, "get"):
-            for choice in response_obj.get("choices", []):
-                response_message = choice.get("message", {})
-                span.set_attribute(
-                    SpanAttributes.OUTPUT_VALUE, response_message.get("content", "")
-                )
 
-                # This shows up under `output_messages` tab on the span page
-                # This code assumes a single response
-                span.set_attribute(
-                    f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_ROLE}",
-                    response_message.get("role"),
-                )
-                span.set_attribute(
-                    f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_CONTENT}",
+        # Captures response tokens, message, and content.
+        if hasattr(response_obj, "get"):
+            for idx, choice in enumerate(response_obj.get("choices", [])):
+                response_message = choice.get("message", {})
+                safe_set_attribute(
+                    span,
+                    SpanAttributes.OUTPUT_VALUE,
                     response_message.get("content", ""),
                 )
 
-            usage = response_obj.get("usage")
+                # This shows up under `output_messages` tab on the span page.
+                prefix = f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.{idx}"
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{MessageAttributes.MESSAGE_ROLE}",
+                    response_message.get("role"),
+                )
+                safe_set_attribute(
+                    span,
+                    f"{prefix}.{MessageAttributes.MESSAGE_CONTENT}",
+                    response_message.get("content", ""),
+                )
+
+            # Token usage info.
+            usage = response_obj and response_obj.get("usage")
             if usage:
-                span.set_attribute(
+                safe_set_attribute(
+                    span,
                     SpanAttributes.LLM_TOKEN_COUNT_TOTAL,
                     usage.get("total_tokens"),
                 )
 
                 # The number of tokens used in the LLM response (completion).
-                span.set_attribute(
+                safe_set_attribute(
+                    span,
                     SpanAttributes.LLM_TOKEN_COUNT_COMPLETION,
                     usage.get("completion_tokens"),
                 )
 
                 # The number of tokens used in the LLM prompt.
-                span.set_attribute(
+                safe_set_attribute(
+                    span,
                     SpanAttributes.LLM_TOKEN_COUNT_PROMPT,
                     usage.get("prompt_tokens"),
                 )
-        pass
+
     except Exception as e:
-        verbose_logger.error(f"Error setting arize attributes: {e}")
+        verbose_logger.error(
+            f"[Arize/Phoenix] Failed to set OpenInference span attributes: {e}"
+        )
+        if hasattr(span, "record_exception"):
+            span.record_exception(e)

litellm/litellm_core_utils/litellm_logging.py

@@ -28,6 +28,7 @@ from litellm._logging import _is_debugging_on, verbose_logger
 from litellm.batches.batch_utils import _handle_completed_batch
 from litellm.caching.caching import DualCache, InMemoryCache
 from litellm.caching.caching_handler import LLMCachingHandler
+
 from litellm.constants import (
     DEFAULT_MOCK_RESPONSE_COMPLETION_TOKEN_COUNT,
     DEFAULT_MOCK_RESPONSE_PROMPT_TOKEN_COUNT,
@@ -36,6 +37,7 @@ from litellm.cost_calculator import (
     RealtimeAPITokenUsageProcessor,
     _select_model_name_for_cost_calc,
 )
+from litellm.integrations.agentops import AgentOps
 from litellm.integrations.anthropic_cache_control_hook import AnthropicCacheControlHook
 from litellm.integrations.arize.arize import ArizeLogger
 from litellm.integrations.custom_guardrail import CustomGuardrail
@@ -2685,7 +2687,15 @@ def _init_custom_logger_compatible_class(  # noqa: PLR0915
     """
     try:
         custom_logger_init_args = custom_logger_init_args or {}
-        if logging_integration == "lago":
+        if logging_integration == "agentops":  # Add AgentOps initialization
+            for callback in _in_memory_loggers:
+                if isinstance(callback, AgentOps):
+                    return callback  # type: ignore
+
+            agentops_logger = AgentOps()
+            _in_memory_loggers.append(agentops_logger)
+            return agentops_logger  # type: ignore
+        elif logging_integration == "lago":
             for callback in _in_memory_loggers:
                 if isinstance(callback, LagoLogger):
                     return callback  # type: ignore

litellm/llms/azure/responses/transformation.py

@@ -1,11 +1,14 @@
-from typing import TYPE_CHECKING, Any, Optional, cast
+from typing import TYPE_CHECKING, Any, Dict, Optional, Tuple, cast
 
 import httpx
 
 import litellm
+from litellm._logging import verbose_logger
 from litellm.llms.openai.responses.transformation import OpenAIResponsesAPIConfig
 from litellm.secret_managers.main import get_secret_str
 from litellm.types.llms.openai import *
+from litellm.types.responses.main import *
+from litellm.types.router import GenericLiteLLMParams
 from litellm.utils import _add_path_to_api_base
 
 if TYPE_CHECKING:
@@ -41,11 +44,7 @@ class AzureOpenAIResponsesAPIConfig(OpenAIResponsesAPIConfig):
     def get_complete_url(
         self,
         api_base: Optional[str],
-        api_key: Optional[str],
-        model: str,
-        optional_params: dict,
         litellm_params: dict,
-        stream: Optional[bool] = None,
     ) -> str:
         """
         Constructs a complete URL for the API request.
@@ -92,3 +91,48 @@ class AzureOpenAIResponsesAPIConfig(OpenAIResponsesAPIConfig):
         final_url = httpx.URL(new_url).copy_with(params=query_params)
 
         return str(final_url)
+
+    #########################################################
+    ########## DELETE RESPONSE API TRANSFORMATION ##############
+    #########################################################
+    def transform_delete_response_api_request(
+        self,
+        response_id: str,
+        api_base: str,
+        litellm_params: GenericLiteLLMParams,
+        headers: dict,
+    ) -> Tuple[str, Dict]:
+        """
+        Transform the delete response API request into a URL and data
+
+        Azure OpenAI API expects the following request:
+        - DELETE /openai/responses/{response_id}?api-version=xxx
+
+        This function handles URLs with query parameters by inserting the response_id
+        at the correct location (before any query parameters).
+        """
+        from urllib.parse import urlparse, urlunparse
+
+        # Parse the URL to separate its components
+        parsed_url = urlparse(api_base)
+
+        # Insert the response_id at the end of the path component
+        # Remove trailing slash if present to avoid double slashes
+        path = parsed_url.path.rstrip("/")
+        new_path = f"{path}/{response_id}"
+
+        # Reconstruct the URL with all original components but with the modified path
+        delete_url = urlunparse(
+            (
+                parsed_url.scheme,  # http, https
+                parsed_url.netloc,  # domain name, port
+                new_path,  # path with response_id added
+                parsed_url.params,  # parameters
+                parsed_url.query,  # query string
+                parsed_url.fragment,  # fragment
+            )
+        )
+
+        data: Dict = {}
+        verbose_logger.debug(f"delete response url={delete_url}")
+        return delete_url, data

litellm/llms/base_llm/responses/transformation.py

@@ -1,6 +1,6 @@
 import types
 from abc import ABC, abstractmethod
-from typing import TYPE_CHECKING, Any, Dict, Optional, Union
+from typing import TYPE_CHECKING, Any, Dict, Optional, Tuple, Union
 
 import httpx
 
@@ -10,6 +10,7 @@ from litellm.types.llms.openai import (
     ResponsesAPIResponse,
     ResponsesAPIStreamingResponse,
 )
+from litellm.types.responses.main import *
 from litellm.types.router import GenericLiteLLMParams
 
 if TYPE_CHECKING:
@@ -73,11 +74,7 @@ class BaseResponsesAPIConfig(ABC):
     def get_complete_url(
         self,
         api_base: Optional[str],
-        api_key: Optional[str],
-        model: str,
-        optional_params: dict,
         litellm_params: dict,
-        stream: Optional[bool] = None,
     ) -> str:
         """
         OPTIONAL
@@ -122,6 +119,31 @@ class BaseResponsesAPIConfig(ABC):
         """
         pass
 
+    #########################################################
+    ########## DELETE RESPONSE API TRANSFORMATION ##############
+    #########################################################
+    @abstractmethod
+    def transform_delete_response_api_request(
+        self,
+        response_id: str,
+        api_base: str,
+        litellm_params: GenericLiteLLMParams,
+        headers: dict,
+    ) -> Tuple[str, Dict]:
+        pass
+
+    @abstractmethod
+    def transform_delete_response_api_response(
+        self,
+        raw_response: httpx.Response,
+        logging_obj: LiteLLMLoggingObj,
+    ) -> DeleteResponseResult:
+        pass
+
+    #########################################################
+    ########## END DELETE RESPONSE API TRANSFORMATION ##########
+    #########################################################
+
     def get_error_class(
         self, error_message: str, status_code: int, headers: Union[dict, httpx.Headers]
     ) -> BaseLLMException:

litellm/llms/custom_httpx/http_handler.py

@@ -650,6 +650,49 @@ class HTTPHandler:
         except Exception as e:
             raise e
 
+    def delete(
+        self,
+        url: str,
+        data: Optional[Union[dict, str]] = None,  # type: ignore
+        json: Optional[dict] = None,
+        params: Optional[dict] = None,
+        headers: Optional[dict] = None,
+        timeout: Optional[Union[float, httpx.Timeout]] = None,
+        stream: bool = False,
+    ):
+        try:
+            if timeout is not None:
+                req = self.client.build_request(
+                    "DELETE", url, data=data, json=json, params=params, headers=headers, timeout=timeout  # type: ignore
+                )
+            else:
+                req = self.client.build_request(
+                    "DELETE", url, data=data, json=json, params=params, headers=headers  # type: ignore
+                )
+            response = self.client.send(req, stream=stream)
+            response.raise_for_status()
+            return response
+        except httpx.TimeoutException:
+            raise litellm.Timeout(
+                message=f"Connection timed out after {timeout} seconds.",
+                model="default-model-name",
+                llm_provider="litellm-httpx-handler",
+            )
+        except httpx.HTTPStatusError as e:
+            if stream is True:
+                setattr(e, "message", mask_sensitive_info(e.response.read()))
+                setattr(e, "text", mask_sensitive_info(e.response.read()))
+            else:
+                error_text = mask_sensitive_info(e.response.text)
+                setattr(e, "message", error_text)
+                setattr(e, "text", error_text)
+
+            setattr(e, "status_code", e.response.status_code)
+
+            raise e
+        except Exception as e:
+            raise e
+
     def __del__(self) -> None:
         try:
             self.close()

litellm/llms/custom_httpx/llm_http_handler.py

@@ -36,6 +36,7 @@ from litellm.types.llms.openai import (
     ResponsesAPIResponse,
 )
 from litellm.types.rerank import OptionalRerankParams, RerankResponse
+from litellm.types.responses.main import DeleteResponseResult
 from litellm.types.router import GenericLiteLLMParams
 from litellm.types.utils import EmbeddingResponse, FileTypes, TranscriptionResponse
 from litellm.utils import CustomStreamWrapper, ModelResponse, ProviderConfigManager
@@ -1015,6 +1016,7 @@ class BaseLLMHTTPHandler:
         client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
         _is_async: bool = False,
         fake_stream: bool = False,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
     ) -> Union[
         ResponsesAPIResponse,
         BaseResponsesAPIStreamingIterator,
@@ -1041,6 +1043,7 @@ class BaseLLMHTTPHandler:
                 timeout=timeout,
                 client=client if isinstance(client, AsyncHTTPHandler) else None,
                 fake_stream=fake_stream,
+                litellm_metadata=litellm_metadata,
             )
 
         if client is None or not isinstance(client, HTTPHandler):
@@ -1064,11 +1067,7 @@ class BaseLLMHTTPHandler:
 
         api_base = responses_api_provider_config.get_complete_url(
             api_base=litellm_params.api_base,
-            api_key=litellm_params.api_key,
-            model=model,
-            optional_params=response_api_optional_request_params,
             litellm_params=dict(litellm_params),
-            stream=stream,
         )
 
         data = responses_api_provider_config.transform_responses_api_request(
@@ -1113,6 +1112,8 @@ class BaseLLMHTTPHandler:
                         model=model,
                         logging_obj=logging_obj,
                         responses_api_provider_config=responses_api_provider_config,
+                        litellm_metadata=litellm_metadata,
+                        custom_llm_provider=custom_llm_provider,
                     )
 
                 return SyncResponsesAPIStreamingIterator(
@@ -1120,6 +1121,8 @@ class BaseLLMHTTPHandler:
                     model=model,
                     logging_obj=logging_obj,
                     responses_api_provider_config=responses_api_provider_config,
+                    litellm_metadata=litellm_metadata,
+                    custom_llm_provider=custom_llm_provider,
                 )
             else:
                 # For non-streaming requests
@@ -1156,6 +1159,7 @@ class BaseLLMHTTPHandler:
         timeout: Optional[Union[float, httpx.Timeout]] = None,
         client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
         fake_stream: bool = False,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
     ) -> Union[ResponsesAPIResponse, BaseResponsesAPIStreamingIterator]:
         """
         Async version of the responses API handler.
@@ -1183,11 +1187,7 @@ class BaseLLMHTTPHandler:
 
         api_base = responses_api_provider_config.get_complete_url(
             api_base=litellm_params.api_base,
-            api_key=litellm_params.api_key,
-            model=model,
-            optional_params=response_api_optional_request_params,
             litellm_params=dict(litellm_params),
-            stream=stream,
         )
 
         data = responses_api_provider_config.transform_responses_api_request(
@@ -1234,6 +1234,8 @@ class BaseLLMHTTPHandler:
                         model=model,
                         logging_obj=logging_obj,
                         responses_api_provider_config=responses_api_provider_config,
+                        litellm_metadata=litellm_metadata,
+                        custom_llm_provider=custom_llm_provider,
                     )
 
                 # Return the streaming iterator
@@ -1242,6 +1244,8 @@ class BaseLLMHTTPHandler:
                     model=model,
                     logging_obj=logging_obj,
                     responses_api_provider_config=responses_api_provider_config,
+                    litellm_metadata=litellm_metadata,
+                    custom_llm_provider=custom_llm_provider,
                 )
             else:
                 # For non-streaming, proceed as before
@@ -1265,6 +1269,163 @@ class BaseLLMHTTPHandler:
             logging_obj=logging_obj,
         )
 
+    async def async_delete_response_api_handler(
+        self,
+        response_id: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        litellm_params: GenericLiteLLMParams,
+        logging_obj: LiteLLMLoggingObj,
+        custom_llm_provider: Optional[str],
+        extra_headers: Optional[Dict[str, Any]] = None,
+        extra_body: Optional[Dict[str, Any]] = None,
+        timeout: Optional[Union[float, httpx.Timeout]] = None,
+        client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
+        _is_async: bool = False,
+    ) -> DeleteResponseResult:
+        """
+        Async version of the delete response API handler.
+        Uses async HTTP client to make requests.
+        """
+        if client is None or not isinstance(client, AsyncHTTPHandler):
+            async_httpx_client = get_async_httpx_client(
+                llm_provider=litellm.LlmProviders(custom_llm_provider),
+                params={"ssl_verify": litellm_params.get("ssl_verify", None)},
+            )
+        else:
+            async_httpx_client = client
+
+        headers = responses_api_provider_config.validate_environment(
+            api_key=litellm_params.api_key,
+            headers=extra_headers or {},
+            model="None",
+        )
+
+        if extra_headers:
+            headers.update(extra_headers)
+
+        api_base = responses_api_provider_config.get_complete_url(
+            api_base=litellm_params.api_base,
+            litellm_params=dict(litellm_params),
+        )
+
+        url, data = responses_api_provider_config.transform_delete_response_api_request(
+            response_id=response_id,
+            api_base=api_base,
+            litellm_params=litellm_params,
+            headers=headers,
+        )
+
+        ## LOGGING
+        logging_obj.pre_call(
+            input=input,
+            api_key="",
+            additional_args={
+                "complete_input_dict": data,
+                "api_base": api_base,
+                "headers": headers,
+            },
+        )
+
+        try:
+            response = await async_httpx_client.delete(
+                url=url, headers=headers, data=json.dumps(data), timeout=timeout
+            )
+
+        except Exception as e:
+            raise self._handle_error(
+                e=e,
+                provider_config=responses_api_provider_config,
+            )
+
+        return responses_api_provider_config.transform_delete_response_api_response(
+            raw_response=response,
+            logging_obj=logging_obj,
+        )
+
+    def delete_response_api_handler(
+        self,
+        response_id: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        litellm_params: GenericLiteLLMParams,
+        logging_obj: LiteLLMLoggingObj,
+        custom_llm_provider: Optional[str],
+        extra_headers: Optional[Dict[str, Any]] = None,
+        extra_body: Optional[Dict[str, Any]] = None,
+        timeout: Optional[Union[float, httpx.Timeout]] = None,
+        client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
+        _is_async: bool = False,
+    ) -> Union[DeleteResponseResult, Coroutine[Any, Any, DeleteResponseResult]]:
+        """
+        Async version of the responses API handler.
+        Uses async HTTP client to make requests.
+        """
+        if _is_async:
+            return self.async_delete_response_api_handler(
+                response_id=response_id,
+                responses_api_provider_config=responses_api_provider_config,
+                litellm_params=litellm_params,
+                logging_obj=logging_obj,
+                custom_llm_provider=custom_llm_provider,
+                extra_headers=extra_headers,
+                extra_body=extra_body,
+                timeout=timeout,
+                client=client,
+            )
+        if client is None or not isinstance(client, HTTPHandler):
+            sync_httpx_client = _get_httpx_client(
+                params={"ssl_verify": litellm_params.get("ssl_verify", None)}
+            )
+        else:
+            sync_httpx_client = client
+
+        headers = responses_api_provider_config.validate_environment(
+            api_key=litellm_params.api_key,
+            headers=extra_headers or {},
+            model="None",
+        )
+
+        if extra_headers:
+            headers.update(extra_headers)
+
+        api_base = responses_api_provider_config.get_complete_url(
+            api_base=litellm_params.api_base,
+            litellm_params=dict(litellm_params),
+        )
+
+        url, data = responses_api_provider_config.transform_delete_response_api_request(
+            response_id=response_id,
+            api_base=api_base,
+            litellm_params=litellm_params,
+            headers=headers,
+        )
+
+        ## LOGGING
+        logging_obj.pre_call(
+            input=input,
+            api_key="",
+            additional_args={
+                "complete_input_dict": data,
+                "api_base": api_base,
+                "headers": headers,
+            },
+        )
+
+        try:
+            response = sync_httpx_client.delete(
+                url=url, headers=headers, data=json.dumps(data), timeout=timeout
+            )
+
+        except Exception as e:
+            raise self._handle_error(
+                e=e,
+                provider_config=responses_api_provider_config,
+            )
+
+        return responses_api_provider_config.transform_delete_response_api_response(
+            raw_response=response,
+            logging_obj=logging_obj,
+        )
+
     def create_file(
         self,
         create_file_data: CreateFileRequest,

litellm/llms/openai/responses/transformation.py

@@ -7,6 +7,7 @@ from litellm._logging import verbose_logger
 from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
 from litellm.secret_managers.main import get_secret_str
 from litellm.types.llms.openai import *
+from litellm.types.responses.main import *
 from litellm.types.router import GenericLiteLLMParams
 
 from ..common_utils import OpenAIError
@@ -110,11 +111,7 @@ class OpenAIResponsesAPIConfig(BaseResponsesAPIConfig):
     def get_complete_url(
         self,
         api_base: Optional[str],
-        api_key: Optional[str],
-        model: str,
-        optional_params: dict,
         litellm_params: dict,
-        stream: Optional[bool] = None,
     ) -> str:
         """
         Get the endpoint for OpenAI responses API
@@ -217,3 +214,39 @@ class OpenAIResponsesAPIConfig(BaseResponsesAPIConfig):
                     f"Error getting model info in OpenAIResponsesAPIConfig: {e}"
                 )
         return False
+
+    #########################################################
+    ########## DELETE RESPONSE API TRANSFORMATION ##############
+    #########################################################
+    def transform_delete_response_api_request(
+        self,
+        response_id: str,
+        api_base: str,
+        litellm_params: GenericLiteLLMParams,
+        headers: dict,
+    ) -> Tuple[str, Dict]:
+        """
+        Transform the delete response API request into a URL and data
+
+        OpenAI API expects the following request
+        - DELETE /v1/responses/{response_id}
+        """
+        url = f"{api_base}/{response_id}"
+        data: Dict = {}
+        return url, data
+
+    def transform_delete_response_api_response(
+        self,
+        raw_response: httpx.Response,
+        logging_obj: LiteLLMLoggingObj,
+    ) -> DeleteResponseResult:
+        """
+        Transform the delete response API response into a DeleteResponseResult
+        """
+        try:
+            raw_response_json = raw_response.json()
+        except Exception:
+            raise OpenAIError(
+                message=raw_response.text, status_code=raw_response.status_code
+            )
+        return DeleteResponseResult(**raw_response_json)

litellm/model_prices_and_context_window_backup.json

@@ -1490,7 +1490,6 @@
         "supports_prompt_caching": false,
         "supports_system_messages": true,
         "supports_tool_choice": true,
-        "supports_native_streaming": false,
         "supports_reasoning": true
     },
     "azure/gpt-4o-audio-preview-2024-12-17": {

litellm/proxy/_types.py

@@ -687,6 +687,8 @@ class GenerateKeyResponse(KeyRequestBase):
     token: Optional[str] = None
     created_by: Optional[str] = None
     updated_by: Optional[str] = None
+    created_at: Optional[datetime] = None
+    updated_at: Optional[datetime] = None
 
     @model_validator(mode="before")
     @classmethod

litellm/proxy/management_endpoints/internal_user_endpoints.py

@@ -43,6 +43,9 @@ from litellm.types.proxy.management_endpoints.common_daily_activity import (
     SpendAnalyticsPaginatedResponse,
     SpendMetrics,
 )
+from litellm.types.proxy.management_endpoints.internal_user_endpoints import (
+    UserListResponse,
+)
 
 router = APIRouter()
 
@@ -899,15 +902,47 @@ async def get_user_key_counts(
     return result
 
 
-@router.get(
-    "/user/get_users",
-    tags=["Internal User management"],
-    dependencies=[Depends(user_api_key_auth)],
-)
+def _validate_sort_params(
+    sort_by: Optional[str], sort_order: str
+) -> Optional[Dict[str, str]]:
+    order_by: Dict[str, str] = {}
+
+    if sort_by is None:
+        return None
+    # Validate sort_by is a valid column
+    valid_columns = [
+        "user_id",
+        "user_email",
+        "created_at",
+        "spend",
+        "user_alias",
+        "user_role",
+    ]
+    if sort_by not in valid_columns:
+        raise HTTPException(
+            status_code=400,
+            detail={
+                "error": f"Invalid sort column. Must be one of: {', '.join(valid_columns)}"
+            },
+        )
+
+    # Validate sort_order
+    if sort_order.lower() not in ["asc", "desc"]:
+        raise HTTPException(
+            status_code=400,
+            detail={"error": "Invalid sort order. Must be 'asc' or 'desc'"},
+        )
+
+    order_by[sort_by] = sort_order.lower()
+
+    return order_by
+
+
 @router.get(
     "/user/list",
     tags=["Internal User management"],
     dependencies=[Depends(user_api_key_auth)],
+    response_model=UserListResponse,
 )
 async def get_users(
     role: Optional[str] = fastapi.Query(
@@ -916,15 +951,29 @@ async def get_users(
     user_ids: Optional[str] = fastapi.Query(
         default=None, description="Get list of users by user_ids"
     ),
+    sso_user_ids: Optional[str] = fastapi.Query(
+        default=None, description="Get list of users by sso_user_id"
+    ),
+    user_email: Optional[str] = fastapi.Query(
+        default=None, description="Filter users by partial email match"
+    ),
+    team: Optional[str] = fastapi.Query(
+        default=None, description="Filter users by team id"
+    ),
     page: int = fastapi.Query(default=1, ge=1, description="Page number"),
     page_size: int = fastapi.Query(
         default=25, ge=1, le=100, description="Number of items per page"
     ),
+    sort_by: Optional[str] = fastapi.Query(
+        default=None,
+        description="Column to sort by (e.g. 'user_id', 'user_email', 'created_at', 'spend')",
+    ),
+    sort_order: str = fastapi.Query(
+        default="asc", description="Sort order ('asc' or 'desc')"
+    ),
 ):
     """
-    Get a paginated list of users, optionally filtered by role.
-
-    Used by the UI to populate the user lists.
+    Get a paginated list of users with filtering and sorting options.
 
     Parameters:
         role: Optional[str]
@@ -935,17 +984,20 @@ async def get_users(
             - internal_user_viewer
         user_ids: Optional[str]
             Get list of users by user_ids. Comma separated list of user_ids.
+        sso_ids: Optional[str]
+            Get list of users by sso_ids. Comma separated list of sso_ids.
+        user_email: Optional[str]
+            Filter users by partial email match
+        team: Optional[str]
+            Filter users by team id. Will match if user has this team in their teams array.
         page: int
             The page number to return
         page_size: int
             The number of items per page
-
-    Currently - admin-only endpoint.
-
-    Example curl:
-    ```
-    http://0.0.0.0:4000/user/list?user_ids=default_user_id,693c1a4a-1cc0-4c7c-afe8-b5d2c8d52e17
-    ```
+        sort_by: Optional[str]
+            Column to sort by (e.g. 'user_id', 'user_email', 'created_at', 'spend')
+        sort_order: Optional[str]
+            Sort order ('asc' or 'desc')
     """
     from litellm.proxy.proxy_server import prisma_client
 
@@ -958,35 +1010,57 @@ async def get_users(
     # Calculate skip and take for pagination
     skip = (page - 1) * page_size
 
-    # Prepare the query conditions
     # Build where conditions based on provided parameters
     where_conditions: Dict[str, Any] = {}
 
     if role:
-        where_conditions["user_role"] = {
-            "contains": role,
-            "mode": "insensitive",  # Case-insensitive search
-        }
+        where_conditions["user_role"] = role  # Exact match instead of contains
 
     if user_ids and isinstance(user_ids, str):
         user_id_list = [uid.strip() for uid in user_ids.split(",") if uid.strip()]
         where_conditions["user_id"] = {
-            "in": user_id_list,  # Now passing a list of strings as required by Prisma
+            "in": user_id_list,
         }
 
-    users: Optional[
-        List[LiteLLM_UserTable]
-    ] = await prisma_client.db.litellm_usertable.find_many(
+    if user_email is not None and isinstance(user_email, str):
+        where_conditions["user_email"] = {
+            "contains": user_email,
+            "mode": "insensitive",  # Case-insensitive search
+        }
+
+    if team is not None and isinstance(team, str):
+        where_conditions["teams"] = {
+            "has": team  # Array contains for string arrays in Prisma
+        }
+
+    if sso_user_ids is not None and isinstance(sso_user_ids, str):
+        sso_id_list = [sid.strip() for sid in sso_user_ids.split(",") if sid.strip()]
+        where_conditions["sso_user_id"] = {
+            "in": sso_id_list,
+        }
+
+    ## Filter any none fastapi.Query params - e.g. where_conditions: {'user_email': {'contains': Query(None), 'mode': 'insensitive'}, 'teams': {'has': Query(None)}}
+    where_conditions = {k: v for k, v in where_conditions.items() if v is not None}
+
+    # Build order_by conditions
+
+    order_by: Optional[Dict[str, str]] = (
+        _validate_sort_params(sort_by, sort_order)
+        if sort_by is not None and isinstance(sort_by, str)
+        else None
+    )
+
+    users = await prisma_client.db.litellm_usertable.find_many(
         where=where_conditions,
         skip=skip,
         take=page_size,
-        order={"created_at": "desc"},
+        order=order_by
+        if order_by
+        else {"created_at": "desc"},  # Default to created_at desc if no sort specified
     )
 
     # Get total count of user rows
-    total_count = await prisma_client.db.litellm_usertable.count(
-        where=where_conditions  # type: ignore
-    )
+    total_count = await prisma_client.db.litellm_usertable.count(where=where_conditions)
 
     # Get key count for each user
     if users is not None:
@@ -1009,7 +1083,7 @@ async def get_users(
                 LiteLLM_UserTableWithKeyCount(
                     **user.model_dump(), key_count=user_key_counts.get(user.user_id, 0)
                 )
-            )  # Return full key object
+            )
     else:
         user_list = []
 

litellm/proxy/management_endpoints/key_management_endpoints.py

@@ -1347,10 +1347,13 @@ async def generate_key_helper_fn(  # noqa: PLR0915
             create_key_response = await prisma_client.insert_data(
                 data=key_data, table_name="key"
             )
+
             key_data["token_id"] = getattr(create_key_response, "token", None)
             key_data["litellm_budget_table"] = getattr(
                 create_key_response, "litellm_budget_table", None
             )
+            key_data["created_at"] = getattr(create_key_response, "created_at", None)
+            key_data["updated_at"] = getattr(create_key_response, "updated_at", None)
     except Exception as e:
         verbose_proxy_logger.error(
             "litellm.proxy.proxy_server.generate_key_helper_fn(): Exception occured - {}".format(

litellm/responses/main.py

@@ -1,7 +1,7 @@
 import asyncio
 import contextvars
 from functools import partial
-from typing import Any, Dict, Iterable, List, Literal, Optional, Union
+from typing import Any, Coroutine, Dict, Iterable, List, Literal, Optional, Union
 
 import httpx
 
@@ -24,6 +24,7 @@ from litellm.types.llms.openai import (
     ToolChoice,
     ToolParam,
 )
+from litellm.types.responses.main import *
 from litellm.types.router import GenericLiteLLMParams
 from litellm.utils import ProviderConfigManager, client
 
@@ -121,7 +122,8 @@ async def aresponses(
         if isinstance(response, ResponsesAPIResponse):
             response = ResponsesAPIRequestUtils._update_responses_api_response_id_with_model_id(
                 responses_api_response=response,
-                kwargs=kwargs,
+                litellm_metadata=kwargs.get("litellm_metadata", {}),
+                custom_llm_provider=custom_llm_provider,
             )
         return response
     except Exception as e:
@@ -253,13 +255,15 @@ def responses(
             fake_stream=responses_api_provider_config.should_fake_stream(
                 model=model, stream=stream, custom_llm_provider=custom_llm_provider
             ),
+            litellm_metadata=kwargs.get("litellm_metadata", {}),
         )
 
         # Update the responses_api_response_id with the model_id
         if isinstance(response, ResponsesAPIResponse):
             response = ResponsesAPIRequestUtils._update_responses_api_response_id_with_model_id(
                 responses_api_response=response,
-                kwargs=kwargs,
+                litellm_metadata=kwargs.get("litellm_metadata", {}),
+                custom_llm_provider=custom_llm_provider,
             )
 
         return response
@@ -271,3 +275,162 @@ def responses(
             completion_kwargs=local_vars,
             extra_kwargs=kwargs,
         )
+
+
+@client
+async def adelete_responses(
+    response_id: str,
+    # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+    # The extra values given here take precedence over values defined on the client or passed to this method.
+    extra_headers: Optional[Dict[str, Any]] = None,
+    extra_query: Optional[Dict[str, Any]] = None,
+    extra_body: Optional[Dict[str, Any]] = None,
+    timeout: Optional[Union[float, httpx.Timeout]] = None,
+    # LiteLLM specific params,
+    custom_llm_provider: Optional[str] = None,
+    **kwargs,
+) -> DeleteResponseResult:
+    """
+    Async version of the DELETE Responses API
+
+    DELETE /v1/responses/{response_id} endpoint in the responses API
+
+    """
+    local_vars = locals()
+    try:
+        loop = asyncio.get_event_loop()
+        kwargs["adelete_responses"] = True
+
+        # get custom llm provider from response_id
+        decoded_response_id: DecodedResponseId = (
+            ResponsesAPIRequestUtils._decode_responses_api_response_id(
+                response_id=response_id,
+            )
+        )
+        response_id = decoded_response_id.get("response_id") or response_id
+        custom_llm_provider = (
+            decoded_response_id.get("custom_llm_provider") or custom_llm_provider
+        )
+
+        func = partial(
+            delete_responses,
+            response_id=response_id,
+            custom_llm_provider=custom_llm_provider,
+            extra_headers=extra_headers,
+            extra_query=extra_query,
+            extra_body=extra_body,
+            timeout=timeout,
+            **kwargs,
+        )
+
+        ctx = contextvars.copy_context()
+        func_with_context = partial(ctx.run, func)
+        init_response = await loop.run_in_executor(None, func_with_context)
+
+        if asyncio.iscoroutine(init_response):
+            response = await init_response
+        else:
+            response = init_response
+        return response
+    except Exception as e:
+        raise litellm.exception_type(
+            model=None,
+            custom_llm_provider=custom_llm_provider,
+            original_exception=e,
+            completion_kwargs=local_vars,
+            extra_kwargs=kwargs,
+        )
+
+
+@client
+def delete_responses(
+    response_id: str,
+    # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+    # The extra values given here take precedence over values defined on the client or passed to this method.
+    extra_headers: Optional[Dict[str, Any]] = None,
+    extra_query: Optional[Dict[str, Any]] = None,
+    extra_body: Optional[Dict[str, Any]] = None,
+    timeout: Optional[Union[float, httpx.Timeout]] = None,
+    # LiteLLM specific params,
+    custom_llm_provider: Optional[str] = None,
+    **kwargs,
+) -> Union[DeleteResponseResult, Coroutine[Any, Any, DeleteResponseResult]]:
+    """
+    Synchronous version of the DELETE Responses API
+
+    DELETE /v1/responses/{response_id} endpoint in the responses API
+
+    """
+    local_vars = locals()
+    try:
+        litellm_logging_obj: LiteLLMLoggingObj = kwargs.get("litellm_logging_obj")  # type: ignore
+        litellm_call_id: Optional[str] = kwargs.get("litellm_call_id", None)
+        _is_async = kwargs.pop("adelete_responses", False) is True
+
+        # get llm provider logic
+        litellm_params = GenericLiteLLMParams(**kwargs)
+
+        # get custom llm provider from response_id
+        decoded_response_id: DecodedResponseId = (
+            ResponsesAPIRequestUtils._decode_responses_api_response_id(
+                response_id=response_id,
+            )
+        )
+        response_id = decoded_response_id.get("response_id") or response_id
+        custom_llm_provider = (
+            decoded_response_id.get("custom_llm_provider") or custom_llm_provider
+        )
+
+        if custom_llm_provider is None:
+            raise ValueError("custom_llm_provider is required but passed as None")
+
+        # get provider config
+        responses_api_provider_config: Optional[BaseResponsesAPIConfig] = (
+            ProviderConfigManager.get_provider_responses_api_config(
+                model=None,
+                provider=litellm.LlmProviders(custom_llm_provider),
+            )
+        )
+
+        if responses_api_provider_config is None:
+            raise ValueError(
+                f"DELETE responses is not supported for {custom_llm_provider}"
+            )
+
+        local_vars.update(kwargs)
+
+        # Pre Call logging
+        litellm_logging_obj.update_environment_variables(
+            model=None,
+            optional_params={
+                "response_id": response_id,
+            },
+            litellm_params={
+                "litellm_call_id": litellm_call_id,
+            },
+            custom_llm_provider=custom_llm_provider,
+        )
+
+        # Call the handler with _is_async flag instead of directly calling the async handler
+        response = base_llm_http_handler.delete_response_api_handler(
+            response_id=response_id,
+            custom_llm_provider=custom_llm_provider,
+            responses_api_provider_config=responses_api_provider_config,
+            litellm_params=litellm_params,
+            logging_obj=litellm_logging_obj,
+            extra_headers=extra_headers,
+            extra_body=extra_body,
+            timeout=timeout or request_timeout,
+            _is_async=_is_async,
+            client=kwargs.get("client"),
+        )
+
+        return response
+    except Exception as e:
+        raise litellm.exception_type(
+            model=None,
+            custom_llm_provider=custom_llm_provider,
+            original_exception=e,
+            completion_kwargs=local_vars,
+            extra_kwargs=kwargs,
+        )

litellm/responses/streaming_iterator.py

@@ -1,7 +1,7 @@
 import asyncio
 import json
 from datetime import datetime
-from typing import Optional
+from typing import Any, Dict, Optional
 
 import httpx
 
@@ -10,6 +10,7 @@ from litellm.litellm_core_utils.asyncify import run_async_function
 from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
 from litellm.litellm_core_utils.thread_pool_executor import executor
 from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
+from litellm.responses.utils import ResponsesAPIRequestUtils
 from litellm.types.llms.openai import (
     OutputTextDeltaEvent,
     ResponseCompletedEvent,
@@ -33,6 +34,8 @@ class BaseResponsesAPIStreamingIterator:
         model: str,
         responses_api_provider_config: BaseResponsesAPIConfig,
         logging_obj: LiteLLMLoggingObj,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
+        custom_llm_provider: Optional[str] = None,
     ):
         self.response = response
         self.model = model
@@ -42,6 +45,10 @@ class BaseResponsesAPIStreamingIterator:
         self.completed_response: Optional[ResponsesAPIStreamingResponse] = None
         self.start_time = datetime.now()
         
+        # set request kwargs
+        self.litellm_metadata = litellm_metadata
+        self.custom_llm_provider = custom_llm_provider
+
     def _process_chunk(self, chunk) -> Optional[ResponsesAPIStreamingResponse]:
         """Process a single chunk of data from the stream"""
         if not chunk:
@@ -70,6 +77,17 @@ class BaseResponsesAPIStreamingIterator:
                         logging_obj=self.logging_obj,
                     )
                 )
+
+                # if "response" in parsed_chunk, then encode litellm specific information like custom_llm_provider
+                response_object = getattr(openai_responses_api_chunk, "response", None)
+                if response_object:
+                    response = ResponsesAPIRequestUtils._update_responses_api_response_id_with_model_id(
+                        responses_api_response=response_object,
+                        litellm_metadata=self.litellm_metadata,
+                        custom_llm_provider=self.custom_llm_provider,
+                    )
+                    setattr(openai_responses_api_chunk, "response", response)
+
                 # Store the completed response
                 if (
                     openai_responses_api_chunk
@@ -102,8 +120,17 @@ class ResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
         model: str,
         responses_api_provider_config: BaseResponsesAPIConfig,
         logging_obj: LiteLLMLoggingObj,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
+        custom_llm_provider: Optional[str] = None,
     ):
-        super().__init__(response, model, responses_api_provider_config, logging_obj)
+        super().__init__(
+            response,
+            model,
+            responses_api_provider_config,
+            logging_obj,
+            litellm_metadata,
+            custom_llm_provider,
+        )
         self.stream_iterator = response.aiter_lines()
 
     def __aiter__(self):
@@ -163,8 +190,17 @@ class SyncResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
         model: str,
         responses_api_provider_config: BaseResponsesAPIConfig,
         logging_obj: LiteLLMLoggingObj,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
+        custom_llm_provider: Optional[str] = None,
     ):
-        super().__init__(response, model, responses_api_provider_config, logging_obj)
+        super().__init__(
+            response,
+            model,
+            responses_api_provider_config,
+            logging_obj,
+            litellm_metadata,
+            custom_llm_provider,
+        )
         self.stream_iterator = response.iter_lines()
 
     def __iter__(self):
@@ -228,12 +264,16 @@ class MockResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
         model: str,
         responses_api_provider_config: BaseResponsesAPIConfig,
         logging_obj: LiteLLMLoggingObj,
+        litellm_metadata: Optional[Dict[str, Any]] = None,
+        custom_llm_provider: Optional[str] = None,
     ):
         super().__init__(
             response=response,
             model=model,
             responses_api_provider_config=responses_api_provider_config,
             logging_obj=logging_obj,
+            litellm_metadata=litellm_metadata,
+            custom_llm_provider=custom_llm_provider,
         )
 
         # one-time transform

litellm/responses/utils.py

@@ -1,5 +1,5 @@
 import base64
-from typing import Any, Dict, Optional, Tuple, Union, cast, get_type_hints
+from typing import Any, Dict, Optional, Union, cast, get_type_hints
 
 import litellm
 from litellm._logging import verbose_logger
@@ -9,6 +9,7 @@ from litellm.types.llms.openai import (
     ResponsesAPIOptionalRequestParams,
     ResponsesAPIResponse,
 )
+from litellm.types.responses.main import DecodedResponseId
 from litellm.types.utils import SpecialEnums, Usage
 
 
@@ -83,30 +84,36 @@ class ResponsesAPIRequestUtils:
     @staticmethod
     def _update_responses_api_response_id_with_model_id(
         responses_api_response: ResponsesAPIResponse,
-        kwargs: Dict[str, Any],
+        custom_llm_provider: Optional[str],
+        litellm_metadata: Optional[Dict[str, Any]] = None,
     ) -> ResponsesAPIResponse:
-        """Update the responses_api_response_id with the model_id"""
-        litellm_metadata: Dict[str, Any] = kwargs.get("litellm_metadata", {}) or {}
+        """
+        Update the responses_api_response_id with model_id and custom_llm_provider
+
+        This builds a composite ID containing the custom LLM provider, model ID, and original response ID
+        """
+        litellm_metadata = litellm_metadata or {}
         model_info: Dict[str, Any] = litellm_metadata.get("model_info", {}) or {}
         model_id = model_info.get("id")
         updated_id = ResponsesAPIRequestUtils._build_responses_api_response_id(
             model_id=model_id,
+            custom_llm_provider=custom_llm_provider,
             response_id=responses_api_response.id,
         )
+
         responses_api_response.id = updated_id
         return responses_api_response
 
     @staticmethod
     def _build_responses_api_response_id(
+        custom_llm_provider: Optional[str],
         model_id: Optional[str],
         response_id: str,
     ) -> str:
         """Build the responses_api_response_id"""
-        if model_id is None:
-            return response_id
         assembled_id: str = str(
             SpecialEnums.LITELLM_MANAGED_RESPONSE_COMPLETE_STR.value
-        ).format(model_id, response_id)
+        ).format(custom_llm_provider, model_id, response_id)
         base64_encoded_id: str = base64.b64encode(assembled_id.encode("utf-8")).decode(
             "utf-8"
         )
@@ -115,12 +122,12 @@ class ResponsesAPIRequestUtils:
     @staticmethod
     def _decode_responses_api_response_id(
         response_id: str,
-    ) -> Tuple[Optional[str], str]:
+    ) -> DecodedResponseId:
         """
         Decode the responses_api_response_id
 
         Returns:
-            Tuple of model_id, response_id (from upstream provider)
+            DecodedResponseId: Structured tuple with custom_llm_provider, model_id, and response_id
         """
         try:
             # Remove prefix and decode
@@ -129,16 +136,45 @@ class ResponsesAPIRequestUtils:
 
             # Parse components using known prefixes
             if ";" not in decoded_id:
-                return None, response_id
+                return DecodedResponseId(
+                    custom_llm_provider=None,
+                    model_id=None,
+                    response_id=response_id,
+                )
 
-            model_part, response_part = decoded_id.split(";", 1)
-            model_id = model_part.replace("litellm:model_id:", "")
+            parts = decoded_id.split(";")
+
+            # Format: litellm:custom_llm_provider:{};model_id:{};response_id:{}
+            custom_llm_provider = None
+            model_id = None
+
+            if (
+                len(parts) >= 3
+            ):  # Full format with custom_llm_provider, model_id, and response_id
+                custom_llm_provider_part = parts[0]
+                model_id_part = parts[1]
+                response_part = parts[2]
+
+                custom_llm_provider = custom_llm_provider_part.replace(
+                    "litellm:custom_llm_provider:", ""
+                )
+                model_id = model_id_part.replace("model_id:", "")
                 decoded_response_id = response_part.replace("response_id:", "")
+            else:
+                decoded_response_id = response_id
 
-            return model_id, decoded_response_id
+            return DecodedResponseId(
+                custom_llm_provider=custom_llm_provider,
+                model_id=model_id,
+                response_id=decoded_response_id,
+            )
         except Exception as e:
             verbose_logger.debug(f"Error decoding response_id '{response_id}': {e}")
-            return None, response_id
+            return DecodedResponseId(
+                custom_llm_provider=None,
+                model_id=None,
+                response_id=response_id,
+            )
 
 
 class ResponseAPILoggingUtils:

litellm/router_utils/pre_call_checks/responses_api_deployment_check.py

@@ -31,11 +31,10 @@ class ResponsesApiDeploymentCheck(CustomLogger):
         if previous_response_id is None:
             return healthy_deployments
 
-        model_id, response_id = (
-            ResponsesAPIRequestUtils._decode_responses_api_response_id(
+        decoded_response = ResponsesAPIRequestUtils._decode_responses_api_response_id(
             response_id=previous_response_id,
         )
-        )
+        model_id = decoded_response.get("model_id")
         if model_id is None:
             return healthy_deployments
 

litellm/types/proxy/management_endpoints/internal_user_endpoints.py

@@ -0,0 +1,18 @@
+from typing import Any, Dict, List, Literal, Optional, Union
+
+from fastapi import HTTPException
+from pydantic import BaseModel, EmailStr
+
+from litellm.proxy._types import LiteLLM_UserTableWithKeyCount
+
+
+class UserListResponse(BaseModel):
+    """
+    Response model for the user list endpoint
+    """
+
+    users: List[LiteLLM_UserTableWithKeyCount]
+    total: int
+    page: int
+    page_size: int
+    total_pages: int

litellm/types/responses/main.py

@@ -1,5 +1,6 @@
 from typing import Literal
 
+from pydantic import PrivateAttr
 from typing_extensions import Any, List, Optional, TypedDict
 
 from litellm.types.llms.base import BaseLiteLLMOpenAIResponseObject
@@ -46,3 +47,30 @@ class GenericResponseOutputItem(BaseLiteLLMOpenAIResponseObject):
     status: str  # "completed", "in_progress", etc.
     role: str  # "assistant", "user", etc.
     content: List[OutputText]
+
+
+class DeleteResponseResult(BaseLiteLLMOpenAIResponseObject):
+    """
+    Result of a delete response request
+
+    {
+        "id": "resp_6786a1bec27481909a17d673315b29f6",
+        "object": "response",
+        "deleted": true
+    }
+    """
+
+    id: Optional[str]
+    object: Optional[str]
+    deleted: Optional[bool]
+
+    # Define private attributes using PrivateAttr
+    _hidden_params: dict = PrivateAttr(default_factory=dict)
+
+
+class DecodedResponseId(TypedDict, total=False):
+    """Structure representing a decoded response ID"""
+
+    custom_llm_provider: Optional[str]
+    model_id: Optional[str]
+    response_id: str

litellm/types/utils.py

@@ -2254,7 +2254,9 @@ class SpecialEnums(Enum):
     LITELM_MANAGED_FILE_ID_PREFIX = "litellm_proxy"
     LITELLM_MANAGED_FILE_COMPLETE_STR = "litellm_proxy:{};unified_id,{}"
 
-    LITELLM_MANAGED_RESPONSE_COMPLETE_STR = "litellm:model_id:{};response_id:{}"
+    LITELLM_MANAGED_RESPONSE_COMPLETE_STR = (
+        "litellm:custom_llm_provider:{};model_id:{};response_id:{}"
+    )
 
 
 LLMResponseTypes = Union[

litellm/utils.py

@@ -180,10 +180,18 @@ from litellm.types.utils import (
     all_litellm_params,
 )
 
-with resources.open_text(
-    "litellm.litellm_core_utils.tokenizers", "anthropic_tokenizer.json"
-) as f:
+try:
+    # Python 3.9+
+    with resources.files("litellm.litellm_core_utils.tokenizers").joinpath(
+        "anthropic_tokenizer.json"
+    ).open("r") as f:
         json_data = json.load(f)
+except (ImportError, AttributeError, TypeError):
+    with resources.open_text(
+        "litellm.litellm_core_utils.tokenizers", "anthropic_tokenizer.json"
+    ) as f:
+        json_data = json.load(f)
+
 # Convert to str (if necessary)
 claude_json_str = json.dumps(json_data)
 import importlib.metadata
@@ -516,9 +524,9 @@ def function_setup(  # noqa: PLR0915
         function_id: Optional[str] = kwargs["id"] if "id" in kwargs else None
 
         ## DYNAMIC CALLBACKS ##
-        dynamic_callbacks: Optional[
-            List[Union[str, Callable, CustomLogger]]
-        ] = kwargs.pop("callbacks", None)
+        dynamic_callbacks: Optional[List[Union[str, Callable, CustomLogger]]] = (
+            kwargs.pop("callbacks", None)
+        )
         all_callbacks = get_dynamic_callbacks(dynamic_callbacks=dynamic_callbacks)
 
         if len(all_callbacks) > 0:
@@ -1202,9 +1210,9 @@ def client(original_function):  # noqa: PLR0915
                         exception=e,
                         retry_policy=kwargs.get("retry_policy"),
                     )
-                    kwargs[
-                        "retry_policy"
-                    ] = reset_retry_policy()  # prevent infinite loops
+                    kwargs["retry_policy"] = (
+                        reset_retry_policy()
+                    )  # prevent infinite loops
                 litellm.num_retries = (
                     None  # set retries to None to prevent infinite loops
                 )
@@ -3028,16 +3036,16 @@ def get_optional_params(  # noqa: PLR0915
                     True  # so that main.py adds the function call to the prompt
                 )
                 if "tools" in non_default_params:
-                    optional_params[
-                        "functions_unsupported_model"
-                    ] = non_default_params.pop("tools")
+                    optional_params["functions_unsupported_model"] = (
+                        non_default_params.pop("tools")
+                    )
                     non_default_params.pop(
                         "tool_choice", None
                     )  # causes ollama requests to hang
                 elif "functions" in non_default_params:
-                    optional_params[
-                        "functions_unsupported_model"
-                    ] = non_default_params.pop("functions")
+                    optional_params["functions_unsupported_model"] = (
+                        non_default_params.pop("functions")
+                    )
             elif (
                 litellm.add_function_to_prompt
             ):  # if user opts to add it to prompt instead
@@ -3060,11 +3068,11 @@ def get_optional_params(  # noqa: PLR0915
 
     if "response_format" in non_default_params:
         if provider_config is not None:
-            non_default_params[
-                "response_format"
-            ] = provider_config.get_json_schema_from_pydantic_object(
+            non_default_params["response_format"] = (
+                provider_config.get_json_schema_from_pydantic_object(
                     response_format=non_default_params["response_format"]
                 )
+            )
         else:
             non_default_params["response_format"] = type_to_response_format_param(
                 response_format=non_default_params["response_format"]
@@ -4079,9 +4087,9 @@ def _count_characters(text: str) -> int:
 
 
 def get_response_string(response_obj: Union[ModelResponse, ModelResponseStream]) -> str:
-    _choices: Union[
-        List[Union[Choices, StreamingChoices]], List[StreamingChoices]
-    ] = response_obj.choices
+    _choices: Union[List[Union[Choices, StreamingChoices]], List[StreamingChoices]] = (
+        response_obj.choices
+    )
 
     response_str = ""
     for choice in _choices:
@@ -6625,8 +6633,8 @@ class ProviderConfigManager:
 
     @staticmethod
     def get_provider_responses_api_config(
-        model: str,
         provider: LlmProviders,
+        model: Optional[str] = None,
     ) -> Optional[BaseResponsesAPIConfig]:
         if litellm.LlmProviders.OPENAI == provider:
             return litellm.OpenAIResponsesAPIConfig()

model_prices_and_context_window.json

@@ -1490,7 +1490,6 @@
         "supports_prompt_caching": false,
         "supports_system_messages": true,
         "supports_tool_choice": true,
-        "supports_native_streaming": false,
         "supports_reasoning": true
     },
     "azure/gpt-4o-audio-preview-2024-12-17": {

poetry.lock

@@ -548,7 +548,7 @@ version = "3.4.1"
 description = "The Real First Universal Charset Detector. Open, modern and actively maintained alternative to Chardet."
 optional = false
 python-versions = ">=3.7"
-groups = ["main"]
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "charset_normalizer-3.4.1-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:91b36a978b5ae0ee86c394f5a54d6ef44db1de0815eb43de826d41d21e4af3de"},
     {file = "charset_normalizer-3.4.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:7461baadb4dc00fd9e0acbe254e3d7d2112e7f92ced2adc96e54ef6501c5f176"},
@@ -742,6 +742,24 @@ ssh = ["bcrypt (>=3.1.5)"]
 test = ["certifi", "cryptography-vectors (==43.0.3)", "pretend", "pytest (>=6.2.0)", "pytest-benchmark", "pytest-cov", "pytest-xdist"]
 test-randomorder = ["pytest-randomly"]
 
+[[package]]
+name = "deprecated"
+version = "1.2.18"
+description = "Python @deprecated decorator to deprecate old python classes, functions or methods."
+optional = false
+python-versions = "!=3.0.*,!=3.1.*,!=3.2.*,!=3.3.*,>=2.7"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "Deprecated-1.2.18-py2.py3-none-any.whl", hash = "sha256:bd5011788200372a32418f888e326a09ff80d0214bd961147cfed01b5c018eec"},
+    {file = "deprecated-1.2.18.tar.gz", hash = "sha256:422b6f6d859da6f2ef57857761bfb392480502a64c3028ca9bbe86085d72115d"},
+]
+
+[package.dependencies]
+wrapt = ">=1.10,<2"
+
+[package.extras]
+dev = ["PyTest", "PyTest-Cov", "bump2version (<1)", "setuptools ; python_version >= \"3.12\"", "tox"]
+
 [[package]]
 name = "distro"
 version = "1.9.0"
@@ -1116,14 +1134,14 @@ protobuf = ">=3.20.2,<4.21.0 || >4.21.0,<4.21.1 || >4.21.1,<4.21.2 || >4.21.2,<4
 name = "googleapis-common-protos"
 version = "1.70.0"
 description = "Common protobufs used in Google APIs"
-optional = true
+optional = false
 python-versions = ">=3.7"
-groups = ["main"]
-markers = "extra == \"extra-proxy\""
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "googleapis_common_protos-1.70.0-py3-none-any.whl", hash = "sha256:b8bfcca8c25a2bb253e0e0b0adaf8c00773e5e6af6fd92397576680b807e0fd8"},
     {file = "googleapis_common_protos-1.70.0.tar.gz", hash = "sha256:0e1b44e0ea153e6594f9f394fef15193a68aaaea2d843f83e2742717ca753257"},
 ]
+markers = {main = "extra == \"extra-proxy\""}
 
 [package.dependencies]
 grpcio = {version = ">=1.44.0,<2.0.0", optional = true, markers = "extra == \"grpc\""}
@@ -1154,10 +1172,9 @@ protobuf = ">=3.20.2,<4.21.1 || >4.21.1,<4.21.2 || >4.21.2,<4.21.3 || >4.21.3,<4
 name = "grpcio"
 version = "1.70.0"
 description = "HTTP/2-based RPC framework"
-optional = true
+optional = false
 python-versions = ">=3.8"
-groups = ["main"]
-markers = "python_version < \"3.10\" and extra == \"extra-proxy\""
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "grpcio-1.70.0-cp310-cp310-linux_armv7l.whl", hash = "sha256:95469d1977429f45fe7df441f586521361e235982a0b39e33841549143ae2851"},
     {file = "grpcio-1.70.0-cp310-cp310-macosx_12_0_universal2.whl", hash = "sha256:ed9718f17fbdb472e33b869c77a16d0b55e166b100ec57b016dc7de9c8d236bf"},
@@ -1215,6 +1232,7 @@ files = [
     {file = "grpcio-1.70.0-cp39-cp39-win_amd64.whl", hash = "sha256:a31d7e3b529c94e930a117b2175b2efd179d96eb3c7a21ccb0289a8ab05b645c"},
     {file = "grpcio-1.70.0.tar.gz", hash = "sha256:8d1584a68d5922330025881e63a6c1b54cc8117291d382e4fa69339b6d914c56"},
 ]
+markers = {main = "python_version < \"3.10\" and extra == \"extra-proxy\"", dev = "python_version < \"3.10\"", proxy-dev = "python_version < \"3.10\""}
 
 [package.extras]
 protobuf = ["grpcio-tools (>=1.70.0)"]
@@ -1223,10 +1241,9 @@ protobuf = ["grpcio-tools (>=1.70.0)"]
 name = "grpcio"
 version = "1.71.0"
 description = "HTTP/2-based RPC framework"
-optional = true
+optional = false
 python-versions = ">=3.9"
-groups = ["main"]
-markers = "python_version >= \"3.10\" and extra == \"extra-proxy\""
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "grpcio-1.71.0-cp310-cp310-linux_armv7l.whl", hash = "sha256:c200cb6f2393468142eb50ab19613229dcc7829b5ccee8b658a36005f6669fdd"},
     {file = "grpcio-1.71.0-cp310-cp310-macosx_12_0_universal2.whl", hash = "sha256:b2266862c5ad664a380fbbcdbdb8289d71464c42a8c29053820ee78ba0119e5d"},
@@ -1280,45 +1297,28 @@ files = [
     {file = "grpcio-1.71.0-cp39-cp39-win_amd64.whl", hash = "sha256:63e41b91032f298b3e973b3fa4093cbbc620c875e2da7b93e249d4728b54559a"},
     {file = "grpcio-1.71.0.tar.gz", hash = "sha256:2b85f7820475ad3edec209d3d89a7909ada16caab05d3f2e08a7e8ae3200a55c"},
 ]
+markers = {main = "python_version >= \"3.10\" and extra == \"extra-proxy\"", dev = "python_version >= \"3.10\"", proxy-dev = "python_version >= \"3.10\""}
 
 [package.extras]
 protobuf = ["grpcio-tools (>=1.71.0)"]
 
 [[package]]
 name = "grpcio-status"
-version = "1.70.0"
+version = "1.62.3"
 description = "Status proto mapping for gRPC"
 optional = true
-python-versions = ">=3.8"
+python-versions = ">=3.6"
 groups = ["main"]
-markers = "python_version < \"3.10\" and extra == \"extra-proxy\""
+markers = "extra == \"extra-proxy\""
 files = [
-    {file = "grpcio_status-1.70.0-py3-none-any.whl", hash = "sha256:fc5a2ae2b9b1c1969cc49f3262676e6854aa2398ec69cb5bd6c47cd501904a85"},
-    {file = "grpcio_status-1.70.0.tar.gz", hash = "sha256:0e7b42816512433b18b9d764285ff029bde059e9d41f8fe10a60631bd8348101"},
+    {file = "grpcio-status-1.62.3.tar.gz", hash = "sha256:289bdd7b2459794a12cf95dc0cb727bd4a1742c37bd823f760236c937e53a485"},
+    {file = "grpcio_status-1.62.3-py3-none-any.whl", hash = "sha256:f9049b762ba8de6b1086789d8315846e094edac2c50beaf462338b301a8fd4b8"},
 ]
 
 [package.dependencies]
 googleapis-common-protos = ">=1.5.5"
-grpcio = ">=1.70.0"
-protobuf = ">=5.26.1,<6.0dev"
-
-[[package]]
-name = "grpcio-status"
-version = "1.71.0"
-description = "Status proto mapping for gRPC"
-optional = true
-python-versions = ">=3.9"
-groups = ["main"]
-markers = "extra == \"extra-proxy\" and python_version >= \"3.10\""
-files = [
-    {file = "grpcio_status-1.71.0-py3-none-any.whl", hash = "sha256:843934ef8c09e3e858952887467f8256aac3910c55f077a359a65b2b3cde3e68"},
-    {file = "grpcio_status-1.71.0.tar.gz", hash = "sha256:11405fed67b68f406b3f3c7c5ae5104a79d2d309666d10d61b152e91d28fb968"},
-]
-
-[package.dependencies]
-googleapis-common-protos = ">=1.5.5"
-grpcio = ">=1.71.0"
-protobuf = ">=5.26.1,<6.0dev"
+grpcio = ">=1.62.3"
+protobuf = ">=4.21.6"
 
 [[package]]
 name = "gunicorn"
@@ -1550,27 +1550,23 @@ all = ["flake8 (>=7.1.1)", "mypy (>=1.11.2)", "pytest (>=8.3.2)", "ruff (>=0.6.2
 
 [[package]]
 name = "importlib-metadata"
-version = "8.5.0"
+version = "7.1.0"
 description = "Read metadata from Python packages"
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev", "proxy-dev"]
 files = [
-    {file = "importlib_metadata-8.5.0-py3-none-any.whl", hash = "sha256:45e54197d28b7a7f1559e60b95e7c567032b602131fbd588f1497f47880aa68b"},
-    {file = "importlib_metadata-8.5.0.tar.gz", hash = "sha256:71522656f0abace1d072b9e5481a48f07c138e00f079c38c8f883823f9c26bd7"},
+    {file = "importlib_metadata-7.1.0-py3-none-any.whl", hash = "sha256:30962b96c0c223483ed6cc7280e7f0199feb01a0e40cfae4d4450fc6fab1f570"},
+    {file = "importlib_metadata-7.1.0.tar.gz", hash = "sha256:b78938b926ee8d5f020fc4772d487045805a55ddbad2ecf21c6d60938dc7fcd2"},
 ]
 
 [package.dependencies]
-zipp = ">=3.20"
+zipp = ">=0.5"
 
 [package.extras]
-check = ["pytest-checkdocs (>=2.4)", "pytest-ruff (>=0.2.1) ; sys_platform != \"cygwin\""]
-cover = ["pytest-cov"]
-doc = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "rst.linker (>=1.9)", "sphinx (>=3.5)", "sphinx-lint"]
-enabler = ["pytest-enabler (>=2.2)"]
+docs = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "rst.linker (>=1.9)", "sphinx (>=3.5)", "sphinx-lint"]
 perf = ["ipython"]
-test = ["flufl.flake8", "importlib-resources (>=1.3) ; python_version < \"3.9\"", "jaraco.test (>=5.4)", "packaging", "pyfakefs", "pytest (>=6,!=8.1.*)", "pytest-perf (>=0.9.2)"]
-type = ["pytest-mypy"]
+testing = ["flufl.flake8", "importlib-resources (>=1.3) ; python_version < \"3.9\"", "jaraco.test (>=5.4)", "packaging", "pyfakefs", "pytest (>=6)", "pytest-checkdocs (>=2.4)", "pytest-cov", "pytest-enabler (>=2.2)", "pytest-mypy ; platform_python_implementation != \"PyPy\"", "pytest-perf (>=0.9.2)", "pytest-ruff (>=0.2.1)"]
 
 [[package]]
 name = "importlib-resources"
@@ -2326,6 +2322,142 @@ datalib = ["numpy (>=1)", "pandas (>=1.2.3)", "pandas-stubs (>=1.1.0.11)"]
 realtime = ["websockets (>=13,<16)"]
 voice-helpers = ["numpy (>=2.0.2)", "sounddevice (>=0.5.1)"]
 
+[[package]]
+name = "opentelemetry-api"
+version = "1.25.0"
+description = "OpenTelemetry Python API"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_api-1.25.0-py3-none-any.whl", hash = "sha256:757fa1aa020a0f8fa139f8959e53dec2051cc26b832e76fa839a6d76ecefd737"},
+    {file = "opentelemetry_api-1.25.0.tar.gz", hash = "sha256:77c4985f62f2614e42ce77ee4c9da5fa5f0bc1e1821085e9a47533a9323ae869"},
+]
+
+[package.dependencies]
+deprecated = ">=1.2.6"
+importlib-metadata = ">=6.0,<=7.1"
+
+[[package]]
+name = "opentelemetry-exporter-otlp"
+version = "1.25.0"
+description = "OpenTelemetry Collector Exporters"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_exporter_otlp-1.25.0-py3-none-any.whl", hash = "sha256:d67a831757014a3bc3174e4cd629ae1493b7ba8d189e8a007003cacb9f1a6b60"},
+    {file = "opentelemetry_exporter_otlp-1.25.0.tar.gz", hash = "sha256:ce03199c1680a845f82e12c0a6a8f61036048c07ec7a0bd943142aca8fa6ced0"},
+]
+
+[package.dependencies]
+opentelemetry-exporter-otlp-proto-grpc = "1.25.0"
+opentelemetry-exporter-otlp-proto-http = "1.25.0"
+
+[[package]]
+name = "opentelemetry-exporter-otlp-proto-common"
+version = "1.25.0"
+description = "OpenTelemetry Protobuf encoding"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_exporter_otlp_proto_common-1.25.0-py3-none-any.whl", hash = "sha256:15637b7d580c2675f70246563363775b4e6de947871e01d0f4e3881d1848d693"},
+    {file = "opentelemetry_exporter_otlp_proto_common-1.25.0.tar.gz", hash = "sha256:c93f4e30da4eee02bacd1e004eb82ce4da143a2f8e15b987a9f603e0a85407d3"},
+]
+
+[package.dependencies]
+opentelemetry-proto = "1.25.0"
+
+[[package]]
+name = "opentelemetry-exporter-otlp-proto-grpc"
+version = "1.25.0"
+description = "OpenTelemetry Collector Protobuf over gRPC Exporter"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_exporter_otlp_proto_grpc-1.25.0-py3-none-any.whl", hash = "sha256:3131028f0c0a155a64c430ca600fd658e8e37043cb13209f0109db5c1a3e4eb4"},
+    {file = "opentelemetry_exporter_otlp_proto_grpc-1.25.0.tar.gz", hash = "sha256:c0b1661415acec5af87625587efa1ccab68b873745ca0ee96b69bb1042087eac"},
+]
+
+[package.dependencies]
+deprecated = ">=1.2.6"
+googleapis-common-protos = ">=1.52,<2.0"
+grpcio = ">=1.0.0,<2.0.0"
+opentelemetry-api = ">=1.15,<2.0"
+opentelemetry-exporter-otlp-proto-common = "1.25.0"
+opentelemetry-proto = "1.25.0"
+opentelemetry-sdk = ">=1.25.0,<1.26.0"
+
+[[package]]
+name = "opentelemetry-exporter-otlp-proto-http"
+version = "1.25.0"
+description = "OpenTelemetry Collector Protobuf over HTTP Exporter"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_exporter_otlp_proto_http-1.25.0-py3-none-any.whl", hash = "sha256:2eca686ee11b27acd28198b3ea5e5863a53d1266b91cda47c839d95d5e0541a6"},
+    {file = "opentelemetry_exporter_otlp_proto_http-1.25.0.tar.gz", hash = "sha256:9f8723859e37c75183ea7afa73a3542f01d0fd274a5b97487ea24cb683d7d684"},
+]
+
+[package.dependencies]
+deprecated = ">=1.2.6"
+googleapis-common-protos = ">=1.52,<2.0"
+opentelemetry-api = ">=1.15,<2.0"
+opentelemetry-exporter-otlp-proto-common = "1.25.0"
+opentelemetry-proto = "1.25.0"
+opentelemetry-sdk = ">=1.25.0,<1.26.0"
+requests = ">=2.7,<3.0"
+
+[[package]]
+name = "opentelemetry-proto"
+version = "1.25.0"
+description = "OpenTelemetry Python Proto"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_proto-1.25.0-py3-none-any.whl", hash = "sha256:f07e3341c78d835d9b86665903b199893befa5e98866f63d22b00d0b7ca4972f"},
+    {file = "opentelemetry_proto-1.25.0.tar.gz", hash = "sha256:35b6ef9dc4a9f7853ecc5006738ad40443701e52c26099e197895cbda8b815a3"},
+]
+
+[package.dependencies]
+protobuf = ">=3.19,<5.0"
+
+[[package]]
+name = "opentelemetry-sdk"
+version = "1.25.0"
+description = "OpenTelemetry Python SDK"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_sdk-1.25.0-py3-none-any.whl", hash = "sha256:d97ff7ec4b351692e9d5a15af570c693b8715ad78b8aafbec5c7100fe966b4c9"},
+    {file = "opentelemetry_sdk-1.25.0.tar.gz", hash = "sha256:ce7fc319c57707ef5bf8b74fb9f8ebdb8bfafbe11898410e0d2a761d08a98ec7"},
+]
+
+[package.dependencies]
+opentelemetry-api = "1.25.0"
+opentelemetry-semantic-conventions = "0.46b0"
+typing-extensions = ">=3.7.4"
+
+[[package]]
+name = "opentelemetry-semantic-conventions"
+version = "0.46b0"
+description = "OpenTelemetry Semantic Conventions"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "opentelemetry_semantic_conventions-0.46b0-py3-none-any.whl", hash = "sha256:6daef4ef9fa51d51855d9f8e0ccd3a1bd59e0e545abe99ac6203804e36ab3e07"},
+    {file = "opentelemetry_semantic_conventions-0.46b0.tar.gz", hash = "sha256:fbc982ecbb6a6e90869b15c1673be90bd18c8a56ff1cffc0864e38e2edffaefa"},
+]
+
+[package.dependencies]
+opentelemetry-api = "1.25.0"
+
 [[package]]
 name = "orjson"
 version = "3.10.15"
@@ -2668,25 +2800,25 @@ testing = ["google-api-core (>=1.31.5)"]
 
 [[package]]
 name = "protobuf"
-version = "5.29.4"
+version = "4.25.6"
 description = ""
-optional = true
+optional = false
 python-versions = ">=3.8"
-groups = ["main"]
-markers = "extra == \"extra-proxy\""
+groups = ["main", "dev", "proxy-dev"]
 files = [
-    {file = "protobuf-5.29.4-cp310-abi3-win32.whl", hash = "sha256:13eb236f8eb9ec34e63fc8b1d6efd2777d062fa6aaa68268fb67cf77f6839ad7"},
-    {file = "protobuf-5.29.4-cp310-abi3-win_amd64.whl", hash = "sha256:bcefcdf3976233f8a502d265eb65ea740c989bacc6c30a58290ed0e519eb4b8d"},
-    {file = "protobuf-5.29.4-cp38-abi3-macosx_10_9_universal2.whl", hash = "sha256:307ecba1d852ec237e9ba668e087326a67564ef83e45a0189a772ede9e854dd0"},
-    {file = "protobuf-5.29.4-cp38-abi3-manylinux2014_aarch64.whl", hash = "sha256:aec4962f9ea93c431d5714ed1be1c93f13e1a8618e70035ba2b0564d9e633f2e"},
-    {file = "protobuf-5.29.4-cp38-abi3-manylinux2014_x86_64.whl", hash = "sha256:d7d3f7d1d5a66ed4942d4fefb12ac4b14a29028b209d4bfb25c68ae172059922"},
-    {file = "protobuf-5.29.4-cp38-cp38-win32.whl", hash = "sha256:1832f0515b62d12d8e6ffc078d7e9eb06969aa6dc13c13e1036e39d73bebc2de"},
-    {file = "protobuf-5.29.4-cp38-cp38-win_amd64.whl", hash = "sha256:476cb7b14914c780605a8cf62e38c2a85f8caff2e28a6a0bad827ec7d6c85d68"},
-    {file = "protobuf-5.29.4-cp39-cp39-win32.whl", hash = "sha256:fd32223020cb25a2cc100366f1dedc904e2d71d9322403224cdde5fdced0dabe"},
-    {file = "protobuf-5.29.4-cp39-cp39-win_amd64.whl", hash = "sha256:678974e1e3a9b975b8bc2447fca458db5f93a2fb6b0c8db46b6675b5b5346812"},
-    {file = "protobuf-5.29.4-py3-none-any.whl", hash = "sha256:3fde11b505e1597f71b875ef2fc52062b6a9740e5f7c8997ce878b6009145862"},
-    {file = "protobuf-5.29.4.tar.gz", hash = "sha256:4f1dfcd7997b31ef8f53ec82781ff434a28bf71d9102ddde14d076adcfc78c99"},
+    {file = "protobuf-4.25.6-cp310-abi3-win32.whl", hash = "sha256:61df6b5786e2b49fc0055f636c1e8f0aff263808bb724b95b164685ac1bcc13a"},
+    {file = "protobuf-4.25.6-cp310-abi3-win_amd64.whl", hash = "sha256:b8f837bfb77513fe0e2f263250f423217a173b6d85135be4d81e96a4653bcd3c"},
+    {file = "protobuf-4.25.6-cp37-abi3-macosx_10_9_universal2.whl", hash = "sha256:6d4381f2417606d7e01750e2729fe6fbcda3f9883aa0c32b51d23012bded6c91"},
+    {file = "protobuf-4.25.6-cp37-abi3-manylinux2014_aarch64.whl", hash = "sha256:5dd800da412ba7f6f26d2c08868a5023ce624e1fdb28bccca2dc957191e81fb5"},
+    {file = "protobuf-4.25.6-cp37-abi3-manylinux2014_x86_64.whl", hash = "sha256:4434ff8bb5576f9e0c78f47c41cdf3a152c0b44de475784cd3fd170aef16205a"},
+    {file = "protobuf-4.25.6-cp38-cp38-win32.whl", hash = "sha256:8bad0f9e8f83c1fbfcc34e573352b17dfce7d0519512df8519994168dc015d7d"},
+    {file = "protobuf-4.25.6-cp38-cp38-win_amd64.whl", hash = "sha256:b6905b68cde3b8243a198268bb46fbec42b3455c88b6b02fb2529d2c306d18fc"},
+    {file = "protobuf-4.25.6-cp39-cp39-win32.whl", hash = "sha256:3f3b0b39db04b509859361ac9bca65a265fe9342e6b9406eda58029f5b1d10b2"},
+    {file = "protobuf-4.25.6-cp39-cp39-win_amd64.whl", hash = "sha256:6ef2045f89d4ad8d95fd43cd84621487832a61d15b49500e4c1350e8a0ef96be"},
+    {file = "protobuf-4.25.6-py3-none-any.whl", hash = "sha256:07972021c8e30b870cfc0863409d033af940213e0e7f64e27fe017b929d2c9f7"},
+    {file = "protobuf-4.25.6.tar.gz", hash = "sha256:f8cfbae7c5afd0d0eaccbe73267339bff605a2315860bb1ba08eb66670a9a91f"},
 ]
+markers = {main = "extra == \"extra-proxy\""}
 
 [[package]]
 name = "pyasn1"
@@ -3342,7 +3474,7 @@ version = "2.31.0"
 description = "Python HTTP for Humans."
 optional = false
 python-versions = ">=3.7"
-groups = ["main"]
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "requests-2.31.0-py3-none-any.whl", hash = "sha256:58cd2187c01e70e6e26505bca751777aa9f2ee0b7f4300988b709f44e013003f"},
     {file = "requests-2.31.0.tar.gz", hash = "sha256:942c5a758f98d790eaed1a29cb6eefc7ffb0d1cf7af05c3d2791656dbd6ad1e1"},
@@ -4027,7 +4159,7 @@ version = "1.26.20"
 description = "HTTP library with thread-safe connection pooling, file post, and more."
 optional = false
 python-versions = "!=3.0.*,!=3.1.*,!=3.2.*,!=3.3.*,!=3.4.*,!=3.5.*,>=2.7"
-groups = ["main"]
+groups = ["main", "dev", "proxy-dev"]
 markers = "python_version < \"3.10\""
 files = [
     {file = "urllib3-1.26.20-py2.py3-none-any.whl", hash = "sha256:0ed14ccfbf1c30a9072c7ca157e4319b70d65f623e91e7b32fadb2853431016e"},
@@ -4045,7 +4177,7 @@ version = "2.2.3"
 description = "HTTP library with thread-safe connection pooling, file post, and more."
 optional = false
 python-versions = ">=3.8"
-groups = ["main", "dev"]
+groups = ["main", "dev", "proxy-dev"]
 markers = "python_version >= \"3.10\""
 files = [
     {file = "urllib3-2.2.3-py3-none-any.whl", hash = "sha256:ca899ca043dcb1bafa3e262d73aa25c465bfb49e0bd9dd5d59f1d0acba2f8fac"},
@@ -4229,6 +4361,95 @@ files = [
     {file = "websockets-13.1.tar.gz", hash = "sha256:a3b3366087c1bc0a2795111edcadddb8b3b59509d5db5d7ea3fdd69f954a8878"},
 ]
 
+[[package]]
+name = "wrapt"
+version = "1.17.2"
+description = "Module for decorators, wrappers and monkey patching."
+optional = false
+python-versions = ">=3.8"
+groups = ["dev", "proxy-dev"]
+files = [
+    {file = "wrapt-1.17.2-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:3d57c572081fed831ad2d26fd430d565b76aa277ed1d30ff4d40670b1c0dd984"},
+    {file = "wrapt-1.17.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:b5e251054542ae57ac7f3fba5d10bfff615b6c2fb09abeb37d2f1463f841ae22"},
+    {file = "wrapt-1.17.2-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:80dd7db6a7cb57ffbc279c4394246414ec99537ae81ffd702443335a61dbf3a7"},
+    {file = "wrapt-1.17.2-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0a6e821770cf99cc586d33833b2ff32faebdbe886bd6322395606cf55153246c"},
+    {file = "wrapt-1.17.2-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:b60fb58b90c6d63779cb0c0c54eeb38941bae3ecf7a73c764c52c88c2dcb9d72"},
+    {file = "wrapt-1.17.2-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b870b5df5b71d8c3359d21be8f0d6c485fa0ebdb6477dda51a1ea54a9b558061"},
+    {file = "wrapt-1.17.2-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:4011d137b9955791f9084749cba9a367c68d50ab8d11d64c50ba1688c9b457f2"},
+    {file = "wrapt-1.17.2-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:1473400e5b2733e58b396a04eb7f35f541e1fb976d0c0724d0223dd607e0f74c"},
+    {file = "wrapt-1.17.2-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:3cedbfa9c940fdad3e6e941db7138e26ce8aad38ab5fe9dcfadfed9db7a54e62"},
+    {file = "wrapt-1.17.2-cp310-cp310-win32.whl", hash = "sha256:582530701bff1dec6779efa00c516496968edd851fba224fbd86e46cc6b73563"},
+    {file = "wrapt-1.17.2-cp310-cp310-win_amd64.whl", hash = "sha256:58705da316756681ad3c9c73fd15499aa4d8c69f9fd38dc8a35e06c12468582f"},
+    {file = "wrapt-1.17.2-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:ff04ef6eec3eee8a5efef2401495967a916feaa353643defcc03fc74fe213b58"},
+    {file = "wrapt-1.17.2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:4db983e7bca53819efdbd64590ee96c9213894272c776966ca6306b73e4affda"},
+    {file = "wrapt-1.17.2-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:9abc77a4ce4c6f2a3168ff34b1da9b0f311a8f1cfd694ec96b0603dff1c79438"},
+    {file = "wrapt-1.17.2-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0b929ac182f5ace000d459c59c2c9c33047e20e935f8e39371fa6e3b85d56f4a"},
+    {file = "wrapt-1.17.2-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f09b286faeff3c750a879d336fb6d8713206fc97af3adc14def0cdd349df6000"},
+    {file = "wrapt-1.17.2-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:1a7ed2d9d039bd41e889f6fb9364554052ca21ce823580f6a07c4ec245c1f5d6"},
+    {file = "wrapt-1.17.2-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:129a150f5c445165ff941fc02ee27df65940fcb8a22a61828b1853c98763a64b"},
+    {file = "wrapt-1.17.2-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:1fb5699e4464afe5c7e65fa51d4f99e0b2eadcc176e4aa33600a3df7801d6662"},
+    {file = "wrapt-1.17.2-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:9a2bce789a5ea90e51a02dfcc39e31b7f1e662bc3317979aa7e5538e3a034f72"},
+    {file = "wrapt-1.17.2-cp311-cp311-win32.whl", hash = "sha256:4afd5814270fdf6380616b321fd31435a462019d834f83c8611a0ce7484c7317"},
+    {file = "wrapt-1.17.2-cp311-cp311-win_amd64.whl", hash = "sha256:acc130bc0375999da18e3d19e5a86403667ac0c4042a094fefb7eec8ebac7cf3"},
+    {file = "wrapt-1.17.2-cp312-cp312-macosx_10_13_universal2.whl", hash = "sha256:d5e2439eecc762cd85e7bd37161d4714aa03a33c5ba884e26c81559817ca0925"},
+    {file = "wrapt-1.17.2-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:3fc7cb4c1c744f8c05cd5f9438a3caa6ab94ce8344e952d7c45a8ed59dd88392"},
+    {file = "wrapt-1.17.2-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:8fdbdb757d5390f7c675e558fd3186d590973244fab0c5fe63d373ade3e99d40"},
+    {file = "wrapt-1.17.2-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5bb1d0dbf99411f3d871deb6faa9aabb9d4e744d67dcaaa05399af89d847a91d"},
+    {file = "wrapt-1.17.2-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d18a4865f46b8579d44e4fe1e2bcbc6472ad83d98e22a26c963d46e4c125ef0b"},
+    {file = "wrapt-1.17.2-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bc570b5f14a79734437cb7b0500376b6b791153314986074486e0b0fa8d71d98"},
+    {file = "wrapt-1.17.2-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:6d9187b01bebc3875bac9b087948a2bccefe464a7d8f627cf6e48b1bbae30f82"},
+    {file = "wrapt-1.17.2-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:9e8659775f1adf02eb1e6f109751268e493c73716ca5761f8acb695e52a756ae"},
+    {file = "wrapt-1.17.2-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:e8b2816ebef96d83657b56306152a93909a83f23994f4b30ad4573b00bd11bb9"},
+    {file = "wrapt-1.17.2-cp312-cp312-win32.whl", hash = "sha256:468090021f391fe0056ad3e807e3d9034e0fd01adcd3bdfba977b6fdf4213ea9"},
+    {file = "wrapt-1.17.2-cp312-cp312-win_amd64.whl", hash = "sha256:ec89ed91f2fa8e3f52ae53cd3cf640d6feff92ba90d62236a81e4e563ac0e991"},
+    {file = "wrapt-1.17.2-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:6ed6ffac43aecfe6d86ec5b74b06a5be33d5bb9243d055141e8cabb12aa08125"},
+    {file = "wrapt-1.17.2-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:35621ae4c00e056adb0009f8e86e28eb4a41a4bfa8f9bfa9fca7d343fe94f998"},
+    {file = "wrapt-1.17.2-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:a604bf7a053f8362d27eb9fefd2097f82600b856d5abe996d623babd067b1ab5"},
+    {file = "wrapt-1.17.2-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5cbabee4f083b6b4cd282f5b817a867cf0b1028c54d445b7ec7cfe6505057cf8"},
+    {file = "wrapt-1.17.2-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:49703ce2ddc220df165bd2962f8e03b84c89fee2d65e1c24a7defff6f988f4d6"},
+    {file = "wrapt-1.17.2-cp313-cp313-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:8112e52c5822fc4253f3901b676c55ddf288614dc7011634e2719718eaa187dc"},
+    {file = "wrapt-1.17.2-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:9fee687dce376205d9a494e9c121e27183b2a3df18037f89d69bd7b35bcf59e2"},
+    {file = "wrapt-1.17.2-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:18983c537e04d11cf027fbb60a1e8dfd5190e2b60cc27bc0808e653e7b218d1b"},
+    {file = "wrapt-1.17.2-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:703919b1633412ab54bcf920ab388735832fdcb9f9a00ae49387f0fe67dad504"},
+    {file = "wrapt-1.17.2-cp313-cp313-win32.whl", hash = "sha256:abbb9e76177c35d4e8568e58650aa6926040d6a9f6f03435b7a522bf1c487f9a"},
+    {file = "wrapt-1.17.2-cp313-cp313-win_amd64.whl", hash = "sha256:69606d7bb691b50a4240ce6b22ebb319c1cfb164e5f6569835058196e0f3a845"},
+    {file = "wrapt-1.17.2-cp313-cp313t-macosx_10_13_universal2.whl", hash = "sha256:4a721d3c943dae44f8e243b380cb645a709ba5bd35d3ad27bc2ed947e9c68192"},
+    {file = "wrapt-1.17.2-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:766d8bbefcb9e00c3ac3b000d9acc51f1b399513f44d77dfe0eb026ad7c9a19b"},
+    {file = "wrapt-1.17.2-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:e496a8ce2c256da1eb98bd15803a79bee00fc351f5dfb9ea82594a3f058309e0"},
+    {file = "wrapt-1.17.2-cp313-cp313t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:40d615e4fe22f4ad3528448c193b218e077656ca9ccb22ce2cb20db730f8d306"},
+    {file = "wrapt-1.17.2-cp313-cp313t-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:a5aaeff38654462bc4b09023918b7f21790efb807f54c000a39d41d69cf552cb"},
+    {file = "wrapt-1.17.2-cp313-cp313t-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:9a7d15bbd2bc99e92e39f49a04653062ee6085c0e18b3b7512a4f2fe91f2d681"},
+    {file = "wrapt-1.17.2-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:e3890b508a23299083e065f435a492b5435eba6e304a7114d2f919d400888cc6"},
+    {file = "wrapt-1.17.2-cp313-cp313t-musllinux_1_2_i686.whl", hash = "sha256:8c8b293cd65ad716d13d8dd3624e42e5a19cc2a2f1acc74b30c2c13f15cb61a6"},
+    {file = "wrapt-1.17.2-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:4c82b8785d98cdd9fed4cac84d765d234ed3251bd6afe34cb7ac523cb93e8b4f"},
+    {file = "wrapt-1.17.2-cp313-cp313t-win32.whl", hash = "sha256:13e6afb7fe71fe7485a4550a8844cc9ffbe263c0f1a1eea569bc7091d4898555"},
+    {file = "wrapt-1.17.2-cp313-cp313t-win_amd64.whl", hash = "sha256:eaf675418ed6b3b31c7a989fd007fa7c3be66ce14e5c3b27336383604c9da85c"},
+    {file = "wrapt-1.17.2-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:5c803c401ea1c1c18de70a06a6f79fcc9c5acfc79133e9869e730ad7f8ad8ef9"},
+    {file = "wrapt-1.17.2-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:f917c1180fdb8623c2b75a99192f4025e412597c50b2ac870f156de8fb101119"},
+    {file = "wrapt-1.17.2-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:ecc840861360ba9d176d413a5489b9a0aff6d6303d7e733e2c4623cfa26904a6"},
+    {file = "wrapt-1.17.2-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:bb87745b2e6dc56361bfde481d5a378dc314b252a98d7dd19a651a3fa58f24a9"},
+    {file = "wrapt-1.17.2-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:58455b79ec2661c3600e65c0a716955adc2410f7383755d537584b0de41b1d8a"},
+    {file = "wrapt-1.17.2-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b4e42a40a5e164cbfdb7b386c966a588b1047558a990981ace551ed7e12ca9c2"},
+    {file = "wrapt-1.17.2-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:91bd7d1773e64019f9288b7a5101f3ae50d3d8e6b1de7edee9c2ccc1d32f0c0a"},
+    {file = "wrapt-1.17.2-cp38-cp38-musllinux_1_2_i686.whl", hash = "sha256:bb90fb8bda722a1b9d48ac1e6c38f923ea757b3baf8ebd0c82e09c5c1a0e7a04"},
+    {file = "wrapt-1.17.2-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:08e7ce672e35efa54c5024936e559469436f8b8096253404faeb54d2a878416f"},
+    {file = "wrapt-1.17.2-cp38-cp38-win32.whl", hash = "sha256:410a92fefd2e0e10d26210e1dfb4a876ddaf8439ef60d6434f21ef8d87efc5b7"},
+    {file = "wrapt-1.17.2-cp38-cp38-win_amd64.whl", hash = "sha256:95c658736ec15602da0ed73f312d410117723914a5c91a14ee4cdd72f1d790b3"},
+    {file = "wrapt-1.17.2-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:99039fa9e6306880572915728d7f6c24a86ec57b0a83f6b2491e1d8ab0235b9a"},
+    {file = "wrapt-1.17.2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:2696993ee1eebd20b8e4ee4356483c4cb696066ddc24bd70bcbb80fa56ff9061"},
+    {file = "wrapt-1.17.2-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:612dff5db80beef9e649c6d803a8d50c409082f1fedc9dbcdfde2983b2025b82"},
+    {file = "wrapt-1.17.2-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:62c2caa1585c82b3f7a7ab56afef7b3602021d6da34fbc1cf234ff139fed3cd9"},
+    {file = "wrapt-1.17.2-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:c958bcfd59bacc2d0249dcfe575e71da54f9dcf4a8bdf89c4cb9a68a1170d73f"},
+    {file = "wrapt-1.17.2-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:fc78a84e2dfbc27afe4b2bd7c80c8db9bca75cc5b85df52bfe634596a1da846b"},
+    {file = "wrapt-1.17.2-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:ba0f0eb61ef00ea10e00eb53a9129501f52385c44853dbd6c4ad3f403603083f"},
+    {file = "wrapt-1.17.2-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:1e1fe0e6ab7775fd842bc39e86f6dcfc4507ab0ffe206093e76d61cde37225c8"},
+    {file = "wrapt-1.17.2-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:c86563182421896d73858e08e1db93afdd2b947a70064b813d515d66549e15f9"},
+    {file = "wrapt-1.17.2-cp39-cp39-win32.whl", hash = "sha256:f393cda562f79828f38a819f4788641ac7c4085f30f1ce1a68672baa686482bb"},
+    {file = "wrapt-1.17.2-cp39-cp39-win_amd64.whl", hash = "sha256:36ccae62f64235cf8ddb682073a60519426fdd4725524ae38874adf72b5f2aeb"},
+    {file = "wrapt-1.17.2-py3-none-any.whl", hash = "sha256:b18f2d1533a71f069c7f82d524a52599053d4c7166e9dd374ae2136b7f40f7c8"},
+    {file = "wrapt-1.17.2.tar.gz", hash = "sha256:41388e9d4d1522446fe79d3213196bd9e3b301a336965b9e27ca2788ebd122f3"},
+]
+
 [[package]]
 name = "wsproto"
 version = "1.2.0"
@@ -4363,7 +4584,7 @@ version = "3.20.2"
 description = "Backport of pathlib-compatible object wrapper for zip files"
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev", "proxy-dev"]
 files = [
     {file = "zipp-3.20.2-py3-none-any.whl", hash = "sha256:a817ac80d6cf4b23bf7f2828b7cabf326f15a001bea8b1f9b49631780ba28350"},
     {file = "zipp-3.20.2.tar.gz", hash = "sha256:bc9eb26f4506fda01b81bcde0ca78103b6e62f991b381fec825435c836edbc29"},
@@ -4384,4 +4605,4 @@ proxy = ["PyJWT", "apscheduler", "backoff", "boto3", "cryptography", "fastapi",
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.8.1,<4.0, !=3.9.7"
-content-hash = "40074b2e47aae8ece058be9a42eda3ca0618e27e4fc9d6529793816df7adb6c8"
+content-hash = "adefc5c35b625ab156ff674c880256643a22880012451d4ade7fa2ef11f5885d"

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "litellm"
-version = "1.67.1"
+version = "1.67.2"
 description = "Library to easily interface with LLM API providers"
 authors = ["BerriAI"]
 license = "MIT"
@@ -107,18 +107,24 @@ types-requests = "*"
 types-setuptools = "*"
 types-redis = "*"
 types-PyYAML = "*"
+opentelemetry-api = "1.25.0"
+opentelemetry-sdk = "1.25.0"
+opentelemetry-exporter-otlp = "1.25.0"
 
 [tool.poetry.group.proxy-dev.dependencies]
 prisma = "0.11.0"
 hypercorn = "^0.15.0"
 prometheus-client = "0.20.0"
+opentelemetry-api = "1.25.0"
+opentelemetry-sdk = "1.25.0"
+opentelemetry-exporter-otlp = "1.25.0"
 
 [build-system]
 requires = ["poetry-core", "wheel"]
 build-backend = "poetry.core.masonry.api"
 
 [tool.commitizen]
-version = "1.67.1"
+version = "1.67.2"
 version_files = [
     "pyproject.toml:^version"
 ]

test-results/.last-run.json

@@ -0,0 +1,4 @@
+{
+  "status": "failed",
+  "failedTests": []
+}

tests/litellm/integrations/arize/test_arize_utils.py

@@ -0,0 +1,231 @@
+import json
+import os
+import sys
+from typing import Optional
+
+# Adds the grandparent directory to sys.path to allow importing project modules
+sys.path.insert(0, os.path.abspath("../.."))
+
+import asyncio
+import litellm
+import pytest
+from litellm.integrations.arize.arize import ArizeLogger
+from litellm.integrations.custom_logger import CustomLogger
+from litellm.integrations._types.open_inference import (
+    SpanAttributes,
+    MessageAttributes,
+    ToolCallAttributes,
+)
+from litellm.types.utils import Choices, StandardCallbackDynamicParams
+
+
+def test_arize_set_attributes():
+    """
+    Test setting attributes for Arize, including all custom LLM attributes.
+    Ensures that the correct span attributes are being added during a request.
+    """
+    from unittest.mock import MagicMock
+    from litellm.types.utils import ModelResponse
+
+    span = MagicMock()  # Mocked tracing span to test attribute setting
+
+    # Construct kwargs to simulate a real LLM request scenario
+    kwargs = {
+        "model": "gpt-4o",
+        "messages": [{"role": "user", "content": "Basic Request Content"}],
+        "standard_logging_object": {
+            "model_parameters": {"user": "test_user"},
+            "metadata": {"key_1": "value_1", "key_2": None},
+            "call_type": "completion",
+        },
+        "optional_params": {
+            "max_tokens": "100",
+            "temperature": "1",
+            "top_p": "5",
+            "stream": False,
+            "user": "test_user",
+            "tools": [
+                {
+                    "function": {
+                        "name": "get_weather",
+                        "description": "Fetches weather details.",
+                        "parameters": {
+                            "type": "object",
+                            "properties": {
+                                "location": {
+                                    "type": "string",
+                                    "description": "City name",
+                                }
+                            },
+                            "required": ["location"],
+                        },
+                    }
+                }
+            ],
+            "functions": [{"name": "get_weather"}, {"name": "get_stock_price"}],
+        },
+        "litellm_params": {"custom_llm_provider": "openai"},
+    }
+
+    # Simulated LLM response object
+    response_obj = ModelResponse(
+        usage={"total_tokens": 100, "completion_tokens": 60, "prompt_tokens": 40},
+        choices=[
+            Choices(message={"role": "assistant", "content": "Basic Response Content"})
+        ],
+        model="gpt-4o",
+        id="chatcmpl-ID",
+    )
+
+    # Apply attribute setting via ArizeLogger
+    ArizeLogger.set_arize_attributes(span, kwargs, response_obj)
+
+    # Validate that the expected number of attributes were set
+    assert span.set_attribute.call_count == 28
+
+    # Metadata attached to the span
+    span.set_attribute.assert_any_call(
+        SpanAttributes.METADATA, json.dumps({"key_1": "value_1", "key_2": None})
+    )
+
+    # Basic LLM information
+    span.set_attribute.assert_any_call(SpanAttributes.LLM_MODEL_NAME, "gpt-4o")
+    span.set_attribute.assert_any_call("llm.request.type", "completion")
+    span.set_attribute.assert_any_call(SpanAttributes.LLM_PROVIDER, "openai")
+
+    # LLM generation parameters
+    span.set_attribute.assert_any_call("llm.request.max_tokens", "100")
+    span.set_attribute.assert_any_call("llm.request.temperature", "1")
+    span.set_attribute.assert_any_call("llm.request.top_p", "5")
+
+    # Streaming and user info
+    span.set_attribute.assert_any_call("llm.is_streaming", "False")
+    span.set_attribute.assert_any_call("llm.user", "test_user")
+
+    # Response metadata
+    span.set_attribute.assert_any_call("llm.response.id", "chatcmpl-ID")
+    span.set_attribute.assert_any_call("llm.response.model", "gpt-4o")
+    span.set_attribute.assert_any_call(SpanAttributes.OPENINFERENCE_SPAN_KIND, "LLM")
+
+    # Request message content and metadata
+    span.set_attribute.assert_any_call(
+        SpanAttributes.INPUT_VALUE, "Basic Request Content"
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_INPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_ROLE}",
+        "user",
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_INPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_CONTENT}",
+        "Basic Request Content",
+    )
+
+    # Tool call definitions and function names
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_TOOLS}.0.{SpanAttributes.TOOL_NAME}", "get_weather"
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_TOOLS}.0.{SpanAttributes.TOOL_DESCRIPTION}",
+        "Fetches weather details.",
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_TOOLS}.0.{SpanAttributes.TOOL_PARAMETERS}",
+        json.dumps(
+            {
+                "type": "object",
+                "properties": {
+                    "location": {"type": "string", "description": "City name"}
+                },
+                "required": ["location"],
+            }
+        ),
+    )
+
+    # Tool calls captured from optional_params
+    span.set_attribute.assert_any_call(
+        f"{MessageAttributes.MESSAGE_TOOL_CALLS}.0.{ToolCallAttributes.TOOL_CALL_FUNCTION_NAME}",
+        "get_weather",
+    )
+    span.set_attribute.assert_any_call(
+        f"{MessageAttributes.MESSAGE_TOOL_CALLS}.1.{ToolCallAttributes.TOOL_CALL_FUNCTION_NAME}",
+        "get_stock_price",
+    )
+
+    # Invocation parameters
+    span.set_attribute.assert_any_call(
+        SpanAttributes.LLM_INVOCATION_PARAMETERS, '{"user": "test_user"}'
+    )
+
+    # User ID
+    span.set_attribute.assert_any_call(SpanAttributes.USER_ID, "test_user")
+
+    # Output message content
+    span.set_attribute.assert_any_call(
+        SpanAttributes.OUTPUT_VALUE, "Basic Response Content"
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_ROLE}",
+        "assistant",
+    )
+    span.set_attribute.assert_any_call(
+        f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_CONTENT}",
+        "Basic Response Content",
+    )
+
+    # Token counts
+    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_TOTAL, 100)
+    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_COMPLETION, 60)
+    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_PROMPT, 40)
+
+
+class TestArizeLogger(CustomLogger):
+    """
+    Custom logger implementation to capture standard_callback_dynamic_params.
+    Used to verify that dynamic config keys are being passed to callbacks.
+    """
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.standard_callback_dynamic_params: Optional[
+            StandardCallbackDynamicParams
+        ] = None
+
+    async def async_log_success_event(self, kwargs, response_obj, start_time, end_time):
+        # Capture dynamic params and print them for verification
+        print("logged kwargs", json.dumps(kwargs, indent=4, default=str))
+        self.standard_callback_dynamic_params = kwargs.get(
+            "standard_callback_dynamic_params"
+        )
+
+
+@pytest.mark.asyncio
+async def test_arize_dynamic_params():
+    """
+    Test to ensure that dynamic Arize keys (API key and space key)
+    are received inside the callback logger at runtime.
+    """
+    test_arize_logger = TestArizeLogger()
+    litellm.callbacks = [test_arize_logger]
+
+    # Perform a mocked async completion call to trigger logging
+    await litellm.acompletion(
+        model="gpt-4o",
+        messages=[{"role": "user", "content": "Basic Request Content"}],
+        mock_response="test",
+        arize_api_key="test_api_key_dynamic",
+        arize_space_key="test_space_key_dynamic",
+    )
+
+    # Allow for async propagation
+    await asyncio.sleep(2)
+
+    # Assert dynamic parameters were received in the callback
+    assert test_arize_logger.standard_callback_dynamic_params is not None
+    assert (
+        test_arize_logger.standard_callback_dynamic_params.get("arize_api_key")
+        == "test_api_key_dynamic"
+    )
+    assert (
+        test_arize_logger.standard_callback_dynamic_params.get("arize_space_key")
+        == "test_space_key_dynamic"
+    )

tests/litellm/integrations/test_agentops.py

@@ -0,0 +1,98 @@
+import os
+import sys
+import pytest
+from unittest.mock import patch, MagicMock
+
+sys.path.insert(0, os.path.abspath("../.."))  # Adds the parent directory to the system-path
+
+from litellm.integrations.agentops.agentops import AgentOps, AgentOpsConfig
+
+@pytest.fixture
+def mock_auth_response():
+    return {
+        "token": "test_jwt_token",
+        "project_id": "test_project_id"
+    }
+
+@pytest.fixture
+def agentops_config():
+    return AgentOpsConfig(
+        endpoint="https://otlp.agentops.cloud/v1/traces",
+        api_key="test_api_key",
+        service_name="test_service",
+        deployment_environment="test_env",
+        auth_endpoint="https://api.agentops.ai/v3/auth/token"
+    )
+
+def test_agentops_config_from_env():
+    """Test that AgentOpsConfig correctly reads from environment variables"""
+    with patch.dict(os.environ, {
+        "AGENTOPS_API_KEY": "test_key",
+        "AGENTOPS_SERVICE_NAME": "test_service",
+        "AGENTOPS_ENVIRONMENT": "test_env"
+    }):
+        config = AgentOpsConfig.from_env()
+        assert config.api_key == "test_key"
+        assert config.service_name == "test_service"
+        assert config.deployment_environment == "test_env"
+        assert config.endpoint == "https://otlp.agentops.cloud/v1/traces"
+        assert config.auth_endpoint == "https://api.agentops.ai/v3/auth/token"
+
+def test_agentops_config_defaults():
+    """Test that AgentOpsConfig uses correct default values"""
+    config = AgentOpsConfig()
+    assert config.service_name is None
+    assert config.deployment_environment is None
+    assert config.api_key is None
+    assert config.endpoint == "https://otlp.agentops.cloud/v1/traces"
+    assert config.auth_endpoint == "https://api.agentops.ai/v3/auth/token"
+
+@patch('litellm.integrations.agentops.agentops.AgentOps._fetch_auth_token')
+def test_fetch_auth_token_success(mock_fetch_auth_token, mock_auth_response):
+    """Test successful JWT token fetch"""
+    mock_fetch_auth_token.return_value = mock_auth_response
+    
+    config = AgentOpsConfig(api_key="test_key")
+    agentops = AgentOps(config=config)
+    
+    mock_fetch_auth_token.assert_called_once_with("test_key", "https://api.agentops.ai/v3/auth/token")
+    assert agentops.resource_attributes.get("project.id") == mock_auth_response.get("project_id")
+
+@patch('litellm.integrations.agentops.agentops.AgentOps._fetch_auth_token')
+def test_fetch_auth_token_failure(mock_fetch_auth_token):
+    """Test failed JWT token fetch"""
+    mock_fetch_auth_token.side_effect = Exception("Failed to fetch auth token: Unauthorized")
+    
+    config = AgentOpsConfig(api_key="test_key")
+    agentops = AgentOps(config=config)
+    
+    mock_fetch_auth_token.assert_called_once()
+    assert "project.id" not in agentops.resource_attributes
+
+@patch('litellm.integrations.agentops.agentops.AgentOps._fetch_auth_token')
+def test_agentops_initialization(mock_fetch_auth_token, agentops_config, mock_auth_response):
+    """Test AgentOps initialization with config"""
+    mock_fetch_auth_token.return_value = mock_auth_response
+    
+    agentops = AgentOps(config=agentops_config)
+    
+    assert agentops.resource_attributes["service.name"] == "test_service"
+    assert agentops.resource_attributes["deployment.environment"] == "test_env"
+    assert agentops.resource_attributes["telemetry.sdk.name"] == "agentops"
+    assert agentops.resource_attributes["project.id"] == "test_project_id"
+
+def test_agentops_initialization_no_auth():
+    """Test AgentOps initialization without authentication"""
+    test_config = AgentOpsConfig(
+        endpoint="https://otlp.agentops.cloud/v1/traces",
+        api_key=None,  # No API key
+        service_name="test_service",
+        deployment_environment="test_env"
+    )
+    
+    agentops = AgentOps(config=test_config)
+    
+    assert agentops.resource_attributes["service.name"] == "test_service"
+    assert agentops.resource_attributes["deployment.environment"] == "test_env"
+    assert agentops.resource_attributes["telemetry.sdk.name"] == "agentops"
+    assert "project.id" not in agentops.resource_attributes 

tests/litellm/llms/openai/responses/test_openai_responses_transformation.py

@@ -203,9 +203,6 @@ class TestOpenAIResponsesAPIConfig:
 
         result = self.config.get_complete_url(
             api_base=api_base,
-            model=self.model,
-            api_key="test_api_key",
-            optional_params={},
             litellm_params={},
         )
 
@@ -215,9 +212,6 @@ class TestOpenAIResponsesAPIConfig:
         with patch("litellm.api_base", "https://litellm-api-base.example.com/v1"):
             result = self.config.get_complete_url(
                 api_base=None,
-                model=self.model,
-                api_key="test_api_key",
-                optional_params={},
                 litellm_params={},
             )
 
@@ -231,9 +225,6 @@ class TestOpenAIResponsesAPIConfig:
             ):
                 result = self.config.get_complete_url(
                     api_base=None,
-                    model=self.model,
-                    api_key="test_api_key",
-                    optional_params={},
                     litellm_params={},
                 )
 
@@ -247,9 +238,6 @@ class TestOpenAIResponsesAPIConfig:
             ):
                 result = self.config.get_complete_url(
                     api_base=None,
-                    model=self.model,
-                    api_key="test_api_key",
-                    optional_params={},
                     litellm_params={},
                 )
 
@@ -260,9 +248,6 @@ class TestOpenAIResponsesAPIConfig:
 
         result = self.config.get_complete_url(
             api_base=api_base,
-            model=self.model,
-            api_key="test_api_key",
-            optional_params={},
             litellm_params={},
         )
 

tests/litellm/proxy/management_endpoints/test_internal_user_endpoints.py

@@ -153,3 +153,19 @@ async def test_get_users_includes_timestamps(mocker):
     assert user_response.created_at == mock_user_data["created_at"]
     assert user_response.updated_at == mock_user_data["updated_at"]
     assert user_response.key_count == 0
+
+
+def test_validate_sort_params():
+    """
+    Test that validate_sort_params returns None if sort_by is None
+    """
+    from litellm.proxy.management_endpoints.internal_user_endpoints import (
+        _validate_sort_params,
+    )
+
+    assert _validate_sort_params(None, "asc") is None
+    assert _validate_sort_params(None, "desc") is None
+    assert _validate_sort_params("user_id", "asc") == {"user_id": "asc"}
+    assert _validate_sort_params("user_id", "desc") == {"user_id": "desc"}
+    with pytest.raises(Exception):
+        _validate_sort_params("user_id", "invalid")

tests/llm_responses_api_testing/base_responses_api.py

@@ -189,6 +189,90 @@ class BaseResponsesAPITest(ABC):
 
 
 
+    @pytest.mark.parametrize("sync_mode", [False, True])
+    @pytest.mark.asyncio
+    async def test_basic_openai_responses_delete_endpoint(self, sync_mode):
+        litellm._turn_on_debug()
+        litellm.set_verbose = True
+        base_completion_call_args = self.get_base_completion_call_args()
+        if sync_mode:
+            response = litellm.responses(
+                input="Basic ping", max_output_tokens=20,
+                **base_completion_call_args
+            )
+
+            # delete the response
+            if isinstance(response, ResponsesAPIResponse):
+                litellm.delete_responses(
+                    response_id=response.id,
+                    **base_completion_call_args
+                )
+            else:
+                raise ValueError("response is not a ResponsesAPIResponse")
+        else:
+            response = await litellm.aresponses(
+                input="Basic ping", max_output_tokens=20,
+                **base_completion_call_args
+            )
+
+            # async delete the response
+            if isinstance(response, ResponsesAPIResponse):
+                await litellm.adelete_responses(
+                    response_id=response.id,
+                    **base_completion_call_args
+                )
+            else:
+                raise ValueError("response is not a ResponsesAPIResponse")
+    
+
+    @pytest.mark.parametrize("sync_mode", [True, False])
+    @pytest.mark.asyncio
+    async def test_basic_openai_responses_streaming_delete_endpoint(self, sync_mode):
+        #litellm._turn_on_debug()
+        #litellm.set_verbose = True
+        base_completion_call_args = self.get_base_completion_call_args()
+        response_id = None
+        if sync_mode:
+            response_id = None
+            response = litellm.responses(
+                input="Basic ping", max_output_tokens=20,
+                stream=True,
+                **base_completion_call_args
+            )
+            for event in response:
+                print("litellm response=", json.dumps(event, indent=4, default=str))
+                if "response" in event:
+                    response_obj = event.get("response")
+                    if response_obj is not None:
+                        response_id = response_obj.get("id")
+            print("got response_id=", response_id)
+
+            # delete the response
+            assert response_id is not None
+            litellm.delete_responses(
+                response_id=response_id,
+                **base_completion_call_args
+            )
+        else:
+            response = await litellm.aresponses(
+                input="Basic ping", max_output_tokens=20,
+                stream=True,
+                **base_completion_call_args
+            )
+            async for event in response:
+                print("litellm response=", json.dumps(event, indent=4, default=str))
+                if "response" in event:
+                    response_obj = event.get("response")
+                    if response_obj is not None:
+                        response_id = response_obj.get("id")
+            print("got response_id=", response_id)
+
+            # delete the response
+            assert response_id is not None
+            await litellm.adelete_responses(
+                response_id=response_id,
+                **base_completion_call_args
+            )
 
 
 

tests/llm_responses_api_testing/test_anthropic_responses_api.py

@@ -30,6 +30,12 @@ class TestAnthropicResponsesAPITest(BaseResponsesAPITest):
             "model": "anthropic/claude-3-5-sonnet-latest",
         }
     
+    async def test_basic_openai_responses_delete_endpoint(self, sync_mode=False):
+        pass
+    
+    async def test_basic_openai_responses_streaming_delete_endpoint(self, sync_mode=False):
+        pass
+
 
 def test_multiturn_tool_calls():
     # Test streaming response with tools for Anthropic

tests/llm_responses_api_testing/test_openai_responses_api.py

@@ -804,7 +804,7 @@ async def test_openai_o1_pro_response_api(sync_mode):
         print("Response:", json.dumps(response, indent=4, default=str))
 
         # Check that the response has the expected structure
-        assert response["id"] == mock_response["id"]
+        assert response["id"] is not None
         assert response["status"] == "incomplete"
         assert response["incomplete_details"].reason == "max_output_tokens"
         assert response["max_output_tokens"] == 20

tests/logging_callback_tests/test_arize_logging.py

@@ -1,111 +0,0 @@
-import os
-import sys
-import time
-from unittest.mock import Mock, patch
-import json
-import opentelemetry.exporter.otlp.proto.grpc.trace_exporter
-from typing import Optional
-
-sys.path.insert(
-    0, os.path.abspath("../..")
-)  # Adds the parent directory to the system-path
-from litellm.integrations._types.open_inference import SpanAttributes
-from litellm.integrations.arize.arize import ArizeConfig, ArizeLogger
-from litellm.integrations.custom_logger import CustomLogger
-from litellm.main import completion
-import litellm
-from litellm.types.utils import Choices, StandardCallbackDynamicParams
-import pytest
-import asyncio
-
-
-def test_arize_set_attributes():
-    """
-    Test setting attributes for Arize
-    """
-    from unittest.mock import MagicMock
-    from litellm.types.utils import ModelResponse
-
-    span = MagicMock()
-    kwargs = {
-        "role": "user",
-        "content": "simple arize test",
-        "model": "gpt-4o",
-        "messages": [{"role": "user", "content": "basic arize test"}],
-        "standard_logging_object": {
-            "model_parameters": {"user": "test_user"},
-            "metadata": {"key": "value", "key2": None},
-        },
-    }
-    response_obj = ModelResponse(
-        usage={"total_tokens": 100, "completion_tokens": 60, "prompt_tokens": 40},
-        choices=[Choices(message={"role": "assistant", "content": "response content"})],
-    )
-
-    ArizeLogger.set_arize_attributes(span, kwargs, response_obj)
-
-    assert span.set_attribute.call_count == 14
-    span.set_attribute.assert_any_call(
-        SpanAttributes.METADATA, json.dumps({"key": "value", "key2": None})
-    )
-    span.set_attribute.assert_any_call(SpanAttributes.LLM_MODEL_NAME, "gpt-4o")
-    span.set_attribute.assert_any_call(SpanAttributes.OPENINFERENCE_SPAN_KIND, "LLM")
-    span.set_attribute.assert_any_call(SpanAttributes.INPUT_VALUE, "basic arize test")
-    span.set_attribute.assert_any_call("llm.input_messages.0.message.role", "user")
-    span.set_attribute.assert_any_call(
-        "llm.input_messages.0.message.content", "basic arize test"
-    )
-    span.set_attribute.assert_any_call(
-        SpanAttributes.LLM_INVOCATION_PARAMETERS, '{"user": "test_user"}'
-    )
-    span.set_attribute.assert_any_call(SpanAttributes.USER_ID, "test_user")
-    span.set_attribute.assert_any_call(SpanAttributes.OUTPUT_VALUE, "response content")
-    span.set_attribute.assert_any_call(
-        "llm.output_messages.0.message.role", "assistant"
-    )
-    span.set_attribute.assert_any_call(
-        "llm.output_messages.0.message.content", "response content"
-    )
-    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_TOTAL, 100)
-    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_COMPLETION, 60)
-    span.set_attribute.assert_any_call(SpanAttributes.LLM_TOKEN_COUNT_PROMPT, 40)
-
-
-class TestArizeLogger(CustomLogger):
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        self.standard_callback_dynamic_params: Optional[
-            StandardCallbackDynamicParams
-        ] = None
-
-    async def async_log_success_event(self, kwargs, response_obj, start_time, end_time):
-        print("logged kwargs", json.dumps(kwargs, indent=4, default=str))
-        self.standard_callback_dynamic_params = kwargs.get(
-            "standard_callback_dynamic_params"
-        )
-
-
-@pytest.mark.asyncio
-async def test_arize_dynamic_params():
-    """verify arize ai dynamic params are recieved by a callback"""
-    test_arize_logger = TestArizeLogger()
-    litellm.callbacks = [test_arize_logger]
-    await litellm.acompletion(
-        model="gpt-4o",
-        messages=[{"role": "user", "content": "basic arize test"}],
-        mock_response="test",
-        arize_api_key="test_api_key_dynamic",
-        arize_space_key="test_space_key_dynamic",
-    )
-
-    await asyncio.sleep(2)
-
-    assert test_arize_logger.standard_callback_dynamic_params is not None
-    assert (
-        test_arize_logger.standard_callback_dynamic_params.get("arize_api_key")
-        == "test_api_key_dynamic"
-    )
-    assert (
-        test_arize_logger.standard_callback_dynamic_params.get("arize_space_key")
-        == "test_space_key_dynamic"
-    )

tests/logging_callback_tests/test_unit_tests_init_callbacks.py

@@ -38,6 +38,7 @@ from litellm.integrations.langfuse.langfuse_prompt_management import (
     LangfusePromptManagement,
 )
 from litellm.integrations.azure_storage.azure_storage import AzureBlobStorageLogger
+from litellm.integrations.agentops import AgentOps
 from litellm.integrations.humanloop import HumanloopLogger
 from litellm.proxy.hooks.dynamic_rate_limiter import _PROXY_DynamicRateLimitHandler
 from unittest.mock import patch
@@ -75,6 +76,7 @@ callback_class_str_to_classType = {
     "pagerduty": PagerDutyAlerting,
     "gcs_pubsub": GcsPubSubLogger,
     "anthropic_cache_control_hook": AnthropicCacheControlHook,
+    "agentops": AgentOps,
 }
 
 expected_env_vars = {

tests/proxy_admin_ui_tests/e2e_ui_tests/search_users.spec.ts

@@ -0,0 +1,214 @@
+/*
+Search Users in Admin UI
+E2E Test for user search functionality
+
+Tests:
+1. Navigate to Internal Users tab
+2. Verify search input exists
+3. Test search functionality
+4. Verify results update
+5. Test filtering by email, user ID, and SSO user ID
+*/
+
+import { test, expect } from "@playwright/test";
+
+test("user search test", async ({ page }) => {
+  // Set a longer timeout for the entire test
+  test.setTimeout(60000);
+
+  // Enable console logging
+  page.on("console", (msg) => console.log("PAGE LOG:", msg.text()));
+
+  // Login first
+  await page.goto("http://localhost:4000/ui");
+  console.log("Navigated to login page");
+
+  // Wait for login form to be visible
+  await page.waitForSelector('input[name="username"]', { timeout: 10000 });
+  console.log("Login form is visible");
+
+  await page.fill('input[name="username"]', "admin");
+  await page.fill('input[name="password"]', "gm");
+  console.log("Filled login credentials");
+
+  const loginButton = page.locator('input[type="submit"]');
+  await expect(loginButton).toBeEnabled();
+  await loginButton.click();
+  console.log("Clicked login button");
+
+  // Wait for navigation to complete and dashboard to load
+  await page.waitForLoadState("networkidle");
+  console.log("Page loaded after login");
+
+  // Take a screenshot for debugging
+  await page.screenshot({ path: "after-login.png" });
+  console.log("Took screenshot after login");
+
+  // Try to find the Internal User tab with more debugging
+  console.log("Looking for Internal User tab...");
+  const internalUserTab = page.locator("span.ant-menu-title-content", {
+    hasText: "Internal User",
+  });
+
+  // Wait for the tab to be visible
+  await internalUserTab.waitFor({ state: "visible", timeout: 10000 });
+  console.log("Internal User tab is visible");
+
+  // Take another screenshot before clicking
+  await page.screenshot({ path: "before-tab-click.png" });
+  console.log("Took screenshot before tab click");
+
+  await internalUserTab.click();
+  console.log("Clicked Internal User tab");
+
+  // Wait for the page to load and table to be visible
+  await page.waitForSelector("tbody tr", { timeout: 30000 });
+  await page.waitForTimeout(2000); // Additional wait for table to stabilize
+  console.log("Table is visible");
+
+  // Take a final screenshot
+  await page.screenshot({ path: "after-tab-click.png" });
+  console.log("Took screenshot after tab click");
+
+  // Verify search input exists
+  const searchInput = page.locator('input[placeholder="Search by email..."]');
+  await expect(searchInput).toBeVisible();
+  console.log("Search input is visible");
+
+  // Test search functionality
+  const initialUserCount = await page.locator("tbody tr").count();
+  console.log(`Initial user count: ${initialUserCount}`);
+
+  // Perform a search
+  const testEmail = "test@";
+  await searchInput.fill(testEmail);
+  console.log("Filled search input");
+
+  // Wait for the debounced search to complete
+  await page.waitForTimeout(500);
+  console.log("Waited for debounce");
+
+  // Wait for the results count to update
+  await page.waitForFunction((initialCount) => {
+    const currentCount = document.querySelectorAll("tbody tr").length;
+    return currentCount !== initialCount;
+  }, initialUserCount);
+  console.log("Results updated");
+
+  const filteredUserCount = await page.locator("tbody tr").count();
+  console.log(`Filtered user count: ${filteredUserCount}`);
+
+  expect(filteredUserCount).toBeDefined();
+
+  // Clear the search
+  await searchInput.clear();
+  console.log("Cleared search");
+
+  await page.waitForTimeout(500);
+  console.log("Waited for debounce after clear");
+
+  await page.waitForFunction((initialCount) => {
+    const currentCount = document.querySelectorAll("tbody tr").length;
+    return currentCount === initialCount;
+  }, initialUserCount);
+  console.log("Results reset");
+
+  const resetUserCount = await page.locator("tbody tr").count();
+  console.log(`Reset user count: ${resetUserCount}`);
+
+  expect(resetUserCount).toBe(initialUserCount);
+});
+
+test("user filter test", async ({ page }) => {
+  // Set a longer timeout for the entire test
+  test.setTimeout(60000);
+
+  // Enable console logging
+  page.on("console", (msg) => console.log("PAGE LOG:", msg.text()));
+
+  // Login first
+  await page.goto("http://localhost:4000/ui");
+  console.log("Navigated to login page");
+
+  // Wait for login form to be visible
+  await page.waitForSelector('input[name="username"]', { timeout: 10000 });
+  console.log("Login form is visible");
+
+  await page.fill('input[name="username"]', "admin");
+  await page.fill('input[name="password"]', "gm");
+  console.log("Filled login credentials");
+
+  const loginButton = page.locator('input[type="submit"]');
+  await expect(loginButton).toBeEnabled();
+  await loginButton.click();
+  console.log("Clicked login button");
+
+  // Wait for navigation to complete and dashboard to load
+  await page.waitForLoadState("networkidle");
+  console.log("Page loaded after login");
+
+  // Navigate to Internal Users tab
+  const internalUserTab = page.locator("span.ant-menu-title-content", {
+    hasText: "Internal User",
+  });
+  await internalUserTab.waitFor({ state: "visible", timeout: 10000 });
+  await internalUserTab.click();
+  console.log("Clicked Internal User tab");
+
+  // Wait for the page to load and table to be visible
+  await page.waitForSelector("tbody tr", { timeout: 30000 });
+  await page.waitForTimeout(2000); // Additional wait for table to stabilize
+  console.log("Table is visible");
+
+  // Get initial user count
+  const initialUserCount = await page.locator("tbody tr").count();
+  console.log(`Initial user count: ${initialUserCount}`);
+
+  // Click the filter button to show additional filters
+  const filterButton = page.getByRole("button", {
+    name: "Filters",
+    exact: true,
+  });
+  await filterButton.click();
+  console.log("Clicked filter button");
+  await page.waitForTimeout(500); // Wait for filters to appear
+
+  // Test user ID filter
+  const userIdInput = page.locator('input[placeholder="Filter by User ID"]');
+  await expect(userIdInput).toBeVisible();
+  console.log("User ID filter is visible");
+
+  await userIdInput.fill("user");
+  console.log("Filled user ID filter");
+  await page.waitForTimeout(1000);
+  const userIdFilteredCount = await page.locator("tbody tr").count();
+  console.log(`User ID filtered count: ${userIdFilteredCount}`);
+  expect(userIdFilteredCount).toBeLessThan(initialUserCount);
+
+  // Clear user ID filter
+  await userIdInput.clear();
+  await page.waitForTimeout(1000);
+  console.log("Cleared user ID filter");
+
+  // Test SSO user ID filter
+  const ssoUserIdInput = page.locator('input[placeholder="Filter by SSO ID"]');
+  await expect(ssoUserIdInput).toBeVisible();
+  console.log("SSO user ID filter is visible");
+
+  await ssoUserIdInput.fill("sso");
+  console.log("Filled SSO user ID filter");
+  await page.waitForTimeout(1000);
+  const ssoUserIdFilteredCount = await page.locator("tbody tr").count();
+  console.log(`SSO user ID filtered count: ${ssoUserIdFilteredCount}`);
+  expect(ssoUserIdFilteredCount).toBeLessThan(initialUserCount);
+
+  // Clear SSO user ID filter
+  await ssoUserIdInput.clear();
+  await page.waitForTimeout(5000);
+  console.log("Cleared SSO user ID filter");
+
+  // Verify count returns to initial after clearing all filters
+  const finalUserCount = await page.locator("tbody tr").count();
+  console.log(`Final user count: ${finalUserCount}`);
+  expect(finalUserCount).toBe(initialUserCount);
+});

tests/proxy_admin_ui_tests/e2e_ui_tests/view_internal_user.spec.ts

@@ -2,45 +2,74 @@
 Test view internal user page
 */
 
-import { test, expect } from '@playwright/test';
+import { test, expect } from "@playwright/test";
 
-test('view internal user page', async ({ page }) => {
+test("view internal user page", async ({ page }) => {
   // Go to the specified URL
-  await page.goto('http://localhost:4000/ui');
+  await page.goto("http://localhost:4000/ui");
 
   // Enter "admin" in the username input field
-  await page.fill('input[name="username"]', 'admin');
+  await page.fill('input[name="username"]', "admin");
 
   // Enter "gm" in the password input field
-  await page.fill('input[name="password"]', 'gm');
+  await page.fill('input[name="password"]', "gm");
 
-  // Optionally, you can add an assertion to verify the login button is enabled
+  // Click the login button
   const loginButton = page.locator('input[type="submit"]');
   await expect(loginButton).toBeEnabled();
-
-  // Optionally, you can click the login button to submit the form
   await loginButton.click();
 
-  const tabElement = page.locator('span.ant-menu-title-content', { hasText: 'Internal User' });
+  // Wait for the Internal User tab and click it
+  const tabElement = page.locator("span.ant-menu-title-content", {
+    hasText: "Internal User",
+  });
   await tabElement.click();
 
-  // try to click on button 
-  // <button class="bg-blue-500 hover:bg-blue-700 text-white font-bold py-2 px-4 rounded-l focus:outline-none" disabled="">← Prev</button>
-  // wait 1-2 seconds
-  await page.waitForTimeout(10000);
+  // Wait for the table to load
+  await page.waitForSelector("tbody tr", { timeout: 10000 });
+  await page.waitForTimeout(2000); // Additional wait for table to stabilize
 
   // Test all expected fields are present
   // number of keys owned by user
-  const keysBadges = page.locator('p.tremor-Badge-text.text-sm.whitespace-nowrap', { hasText: 'Keys' });
-  const keysCountArray = await keysBadges.evaluateAll(elements => elements.map(el => parseInt(el.textContent.split(' ')[0], 10)));
+  const keysBadges = page.locator(
+    "p.tremor-Badge-text.text-sm.whitespace-nowrap",
+    { hasText: "Keys" }
+  );
+  const keysCountArray = await keysBadges.evaluateAll((elements) =>
+    elements.map((el) => {
+      const text = el.textContent;
+      return text ? parseInt(text.split(" ")[0], 10) : 0;
+    })
+  );
 
-  const hasNonZeroKeys = keysCountArray.some(count => count > 0);
+  const hasNonZeroKeys = keysCountArray.some((count) => count > 0);
   expect(hasNonZeroKeys).toBe(true);
 
   // test pagination
-  const prevButton = page.locator('button.px-3.py-1.text-sm.border.rounded-md.hover\\:bg-gray-50.disabled\\:opacity-50.disabled\\:cursor-not-allowed', { hasText: 'Previous' });
-  await expect(prevButton).toBeDisabled();
+  // Wait for pagination controls to be visible
+  await page.waitForSelector(".flex.justify-between.items-center", {
+    timeout: 5000,
+  });
 
-  const nextButton = page.locator('button.px-3.py-1.text-sm.border.rounded-md.hover\\:bg-gray-50.disabled\\:opacity-50.disabled\\:cursor-not-allowed', { hasText: 'Next' });
+  // Check if we're on the first page by looking at the results count
+  const resultsText =
+    (await page.locator(".text-sm.text-gray-700").textContent()) || "";
+  const isFirstPage = resultsText.includes("1 -");
+
+  if (isFirstPage) {
+    // On first page, previous button should be disabled
+    const prevButton = page.locator("button", { hasText: "Previous" });
+    await expect(prevButton).toBeDisabled();
+  }
+
+  // Next button should be enabled if there are more pages
+  const nextButton = page.locator("button", { hasText: "Next" });
+  const totalResults =
+    (await page.locator(".text-sm.text-gray-700").textContent()) || "";
+  const hasMorePages =
+    totalResults.includes("of") && !totalResults.includes("1 - 25 of 25");
+
+  if (hasMorePages) {
     await expect(nextButton).toBeEnabled();
+  }
 });

tests/proxy_admin_ui_tests/e2e_ui_tests/view_user_info.spec.ts

@@ -0,0 +1,82 @@
+import { test, expect } from "@playwright/test";
+import { loginToUI } from "../utils/login";
+
+test.describe("User Info View", () => {
+  test.beforeEach(async ({ page }) => {
+    await loginToUI(page);
+    // Navigate to users page
+    await page.goto("http://localhost:4000/ui?page=users");
+  });
+
+  test("should display user info when clicking on user ID", async ({
+    page,
+  }) => {
+    // Wait for users table to load
+    await page.waitForSelector("table");
+
+    // Get the first user ID cell
+    const firstUserIdCell = page.locator(
+      "table tbody tr:first-child td:first-child"
+    );
+    const userId = await firstUserIdCell.textContent();
+    console.log("Found user ID:", userId);
+
+    // Click on the user ID
+    await firstUserIdCell.click();
+
+    // Wait for user info view to load
+    await page.waitForSelector('h1:has-text("User")');
+    console.log("User info view loaded");
+
+    // Check for tabs
+    await expect(page.locator('button:has-text("Overview")')).toBeVisible();
+    await expect(page.locator('button:has-text("Details")')).toBeVisible();
+
+    // Switch to details tab
+    await page.locator('button:has-text("Details")').click();
+
+    // Check details section
+    await expect(page.locator("text=User ID")).toBeVisible();
+    await expect(page.locator("text=Email")).toBeVisible();
+
+    // Go back to users list
+    await page.locator('button:has-text("Back to Users")').click();
+
+    // Verify we're back on the users page
+    await expect(page.locator('h1:has-text("Users")')).toBeVisible();
+  });
+
+  // test("should handle user deletion", async ({ page }) => {
+  //   // Wait for users table to load
+  //   await page.waitForSelector("table");
+
+  //   // Get the first user ID cell
+  //   const firstUserIdCell = page.locator(
+  //     "table tbody tr:first-child td:first-child"
+  //   );
+  //   const userId = await firstUserIdCell.textContent();
+
+  //   // Click on the user ID
+  //   await firstUserIdCell.click();
+
+  //   // Wait for user info view to load
+  //   await page.waitForSelector('h1:has-text("User")');
+
+  //   // Click delete button
+  //   await page.locator('button:has-text("Delete User")').click();
+
+  //   // Confirm deletion in modal
+  //   await page.locator('button:has-text("Delete")').click();
+
+  //   // Verify success message
+  //   await expect(page.locator("text=User deleted successfully")).toBeVisible();
+
+  //   // Verify we're back on the users page
+  //   await expect(page.locator('h1:has-text("Users")')).toBeVisible();
+
+  //   // Verify user is no longer in the table
+  //   if (userId) {
+  //     await expect(page.locator(`text=${userId}`)).not.toBeVisible();
+  //   }
+  // });
+});

tests/proxy_admin_ui_tests/test-results/.last-run.json

@@ -0,0 +1,8 @@
+{
+  "status": "failed",
+  "failedTests": [
+    "8306bf902634636ae770-183086b993a71bc98dd6",
+    "1bfc70f64c2dd4741dbb-58cd256736ebe53a2d97",
+    "ea1c46def20befad7a54-cb6c473c41474485b610"
+  ]
+}

tests/proxy_admin_ui_tests/utils/login.ts

@@ -0,0 +1,23 @@
+import { Page, expect } from "@playwright/test";
+
+export async function loginToUI(page: Page) {
+  // Login first
+  await page.goto("http://localhost:4000/ui");
+  console.log("Navigated to login page");
+
+  // Wait for login form to be visible
+  await page.waitForSelector('input[name="username"]', { timeout: 10000 });
+  console.log("Login form is visible");
+
+  await page.fill('input[name="username"]', "admin");
+  await page.fill('input[name="password"]', "gm");
+  console.log("Filled login credentials");
+
+  const loginButton = page.locator('input[type="submit"]');
+  await expect(loginButton).toBeEnabled();
+  await loginButton.click();
+  console.log("Clicked login button");
+
+  // Wait for navigation to complete
+  await page.waitForURL("**/*");
+}

tests/test_keys.py

@@ -109,6 +109,18 @@ async def test_key_gen():
         await asyncio.gather(*tasks)
 
 
+@pytest.mark.asyncio
+async def test_simple_key_gen():
+    async with aiohttp.ClientSession() as session:
+        key_data = await generate_key(session, i=0)
+        key = key_data["key"]
+        assert key_data["token"] is not None
+        assert key_data["token"] != key
+        assert key_data["token_id"] is not None
+        assert key_data["created_at"] is not None
+        assert key_data["updated_at"] is not None
+
+
 @pytest.mark.asyncio
 async def test_key_gen_bad_key():
     """

ui/litellm-dashboard/src/components/networking.tsx

@@ -676,6 +676,12 @@ export const userListCall = async (
   userIDs: string[] | null = null,
   page: number | null = null,
   page_size: number | null = null,
+  userEmail: string | null = null,
+  userRole: string | null = null,
+  team: string | null = null,
+  sso_user_id: string | null = null,
+  sortBy: string | null = null,
+  sortOrder: 'asc' | 'desc' | null = null,
 ) => {
   /**
    * Get all available teams on proxy
@@ -699,6 +705,30 @@ export const userListCall = async (
       queryParams.append('page_size', page_size.toString());
     }
 
+    if (userEmail) {
+      queryParams.append('user_email', userEmail);
+    }
+
+    if (userRole) {
+      queryParams.append('role', userRole);
+    }
+
+    if (team) {
+      queryParams.append('team', team);
+    }
+
+    if (sso_user_id) {
+      queryParams.append('sso_user_ids', sso_user_id);
+    }
+
+    if (sortBy) {
+      queryParams.append('sort_by', sortBy);
+    }
+
+    if (sortOrder) {
+      queryParams.append('sort_order', sortOrder);
+    }
+    
     const queryString = queryParams.toString();
     if (queryString) {
       url += `?${queryString}`;
@@ -736,8 +766,10 @@ export const userInfoCall = async (
   userRole: String,
   viewAll: Boolean = false,
   page: number | null,
-  page_size: number | null
+  page_size: number | null,
+  lookup_user_id: boolean = false
 ) => {
+  console.log(`userInfoCall: ${userID}, ${userRole}, ${viewAll}, ${page}, ${page_size}, ${lookup_user_id}`)
   try {
     let url: string;
     
@@ -751,7 +783,7 @@ export const userInfoCall = async (
     } else {
       // Use /user/info endpoint for individual user info
       url = proxyBaseUrl ? `${proxyBaseUrl}/user/info` : `/user/info`;
-      if (userRole === "Admin" || userRole === "Admin Viewer") {
+      if ((userRole === "Admin" || userRole === "Admin Viewer") && !lookup_user_id) {
         // do nothing 
       } else if (userID) {
         url += `?user_id=${userID}`;

ui/litellm-dashboard/src/components/view_users.tsx

@@ -1,4 +1,4 @@
-import React, { useState, useEffect } from "react";
+import React, { useState, useEffect, useCallback, useRef } from "react";
 import {
   Card,
   Title,
@@ -23,6 +23,7 @@ import {
   DialogPanel,
   Icon,
   TextInput,
+  NumberInput,
 } from "@tremor/react";
 
 import { message } from "antd";
@@ -32,6 +33,7 @@ import {
   userInfoCall,
   userUpdateUserCall,
   getPossibleUserRoles,
+  userListCall,
 } from "./networking";
 import { Badge, BadgeDelta, Button } from "@tremor/react";
 import RequestAccess from "./request_model_access";
@@ -50,6 +52,7 @@ import { UserDataTable } from "./view_users/table";
 import { UserInfo } from "./view_users/types";
 import BulkCreateUsers from "./bulk_create_users_button";
 import SSOSettings from "./SSOSettings";
+import debounce from "lodash/debounce";
 
 interface ViewUserDashboardProps {
   accessToken: string | null;
@@ -77,6 +80,19 @@ interface CreateuserProps {
   onUserCreated: () => Promise<void>;
 }
 
+interface FilterState {
+  email: string;
+  user_id: string;
+  user_role: string;
+  sso_user_id: string;
+  team: string;
+  model: string;
+  min_spend: number | null;
+  max_spend: number | null;
+  sort_by: string;
+  sort_order: 'asc' | 'desc';
+}
+
 const isLocal = process.env.NODE_ENV === "development";
 const proxyBaseUrl = isLocal ? "http://localhost:4000" : null;
 if (isLocal != true) {
@@ -93,7 +109,6 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
   setKeys,
 }) => {
   const [userListResponse, setUserListResponse] = useState<UserListResponse | null>(null);
-  const [userData, setUserData] = useState<null | any[]>(null);
   const [endUsers, setEndUsers] = useState<null | any[]>(null);
   const [currentPage, setCurrentPage] = useState(1);
   const [openDialogId, setOpenDialogId] = React.useState<null | number>(null);
@@ -108,6 +123,23 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
   const defaultPageSize = 25;
   const [searchTerm, setSearchTerm] = useState("");
   const [activeTab, setActiveTab] = useState("users");
+  const [filters, setFilters] = useState<FilterState>({
+    email: "",
+    user_id: "",
+    user_role: "",
+    sso_user_id: "",
+    team: "",
+    model: "",
+    min_spend: null,
+    max_spend: null,
+    sort_by: "created_at",
+    sort_order: "desc"
+  });
+  const [showFilters, setShowFilters] = useState(false);
+  const [showColumnDropdown, setShowColumnDropdown] = useState(false);
+  const [selectedFilter, setSelectedFilter] = useState("Email");
+  const filtersRef = useRef(null);
+  const lastSearchTimestamp = useRef(0);
 
   // check if window is not undefined
   if (typeof window !== "undefined") {
@@ -122,15 +154,88 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
     setIsDeleteModalOpen(true);
   };
 
+  const handleFilterChange = (key: keyof FilterState, value: string | number | null) => {
+    const newFilters = { ...filters, [key]: value };
+    setFilters(newFilters);
+    console.log("called from handleFilterChange - newFilters:", JSON.stringify(newFilters));
+    debouncedSearch(newFilters);
+  };
+
+  const handleSortChange = (sortBy: string, sortOrder: 'asc' | 'desc') => {
+    const newFilters = {
+      ...filters,
+      sort_by: sortBy,
+      sort_order: sortOrder
+    };
+    setFilters(newFilters);
+    debouncedSearch(newFilters);
+  };
+
+  // Create a debounced version of the search function
+  const debouncedSearch = useCallback(
+    debounce(async (filters: FilterState) => {
+      if (!accessToken || !token || !userRole || !userID) {
+        return;
+      }
+
+      const currentTimestamp = Date.now();
+      lastSearchTimestamp.current = currentTimestamp;
+
+      try {
+        // Make the API call using userListCall with all filter parameters
+        const data = await userListCall(
+          accessToken,
+          filters.user_id ? [filters.user_id] : null,
+          1, // Reset to first page when searching
+          defaultPageSize,
+          filters.email || null,
+          filters.user_role || null,
+          filters.team || null,
+          filters.sso_user_id || null,
+          filters.sort_by,
+          filters.sort_order
+        );
+        
+        // Only update state if this is the most recent search
+        if (currentTimestamp === lastSearchTimestamp.current) {
+          if (data) {
+            setUserListResponse(data);
+            console.log("called from debouncedSearch filters:", JSON.stringify(filters));
+            console.log("called from debouncedSearch data:", JSON.stringify(data));
+          }
+        }
+      } catch (error) {
+        console.error("Error searching users:", error);
+      }
+    }, 300),
+    [accessToken, token, userRole, userID]
+  );
+
+  // Cleanup the debounced function on component unmount
+  useEffect(() => {
+    return () => {
+      debouncedSearch.cancel();
+    };
+  }, [debouncedSearch]);
+
+  const handleSearch = (value: string) => {
+    setSearchTerm(value);
+    if (value === "") {
+      refreshUserData(); // Reset to original data when search is cleared
+    } else {
+      debouncedSearch(filters);
+    }
+  };
+
   const confirmDelete = async () => {
     if (userToDelete && accessToken) {
       try {
         await userDeleteCall(accessToken, [userToDelete]);
         message.success("User deleted successfully");
         // Update the user list after deletion
-        if (userData) {
-          const updatedUserData = userData.filter(user => user.user_id !== userToDelete);
-          setUserData(updatedUserData);
+        if (userListResponse) {
+          const updatedUserData = userListResponse.users?.filter(user => user.user_id !== userToDelete);
+          setUserListResponse({ ...userListResponse, users: updatedUserData || [] });
         }
       } catch (error) {
         console.error("Error deleting user:", error);
@@ -164,11 +269,11 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
     } catch (error) {
       console.error("There was an error updating the user", error);
     }
-    if (userData) {
-      const updatedUserData = userData.map((user) =>
+    if (userListResponse) {
+      const updatedUserData = userListResponse.users?.map((user) =>
         user.user_id === editedUser.user_id ? editedUser : user
       );
-      setUserData(updatedUserData);
+      setUserListResponse({ ...userListResponse, users: updatedUserData || [] });
     }
     setSelectedUser(null);
     setEditModalVisible(false);
@@ -176,6 +281,7 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
   };
 
   const refreshUserData = async () => {
+    console.log("called from refreshUserData");
     if (!accessToken || !token || !userRole || !userID) {
       return;
     }
@@ -195,14 +301,45 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
         `userList_${currentPage}`,
         JSON.stringify(userDataResponse)
       );
-      
+      console.log("called from refreshUserData");
       setUserListResponse(userDataResponse);
-      setUserData(userDataResponse.users || []);
     } catch (error) {
       console.error("Error refreshing user data:", error);
     }
   };
 
+  const handlePageChange = async (newPage: number) => {
+    if (!accessToken || !token || !userRole || !userID) {
+      return;
+    }
+    
+    try {
+      const userDataResponse = await userListCall(
+        accessToken,
+        filters.user_id ? [filters.user_id] : null,
+        newPage,
+        defaultPageSize,
+        filters.email || null,
+        filters.user_role || null,
+        filters.team || null,
+        filters.sso_user_id || null,
+        filters.sort_by,
+        filters.sort_order
+      );
+      
+      // Update session storage with new data
+      sessionStorage.setItem(
+        `userList_${newPage}`,
+        JSON.stringify(userDataResponse)
+      );
+      
+      setUserListResponse(userDataResponse);
+      setCurrentPage(newPage);
+    } catch (error) {
+      console.error("Error changing page:", error);
+    }
+  };
+
   useEffect(() => {
     if (!accessToken || !token || !userRole || !userID) {
       return;
@@ -214,16 +351,20 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
         if (cachedUserData) {
           const parsedData = JSON.parse(cachedUserData);
           setUserListResponse(parsedData);
-          setUserData(parsedData.users || []);
+          console.log("called from useEffect");
         } else {
-          // Fetch from API if not in cache
-          const userDataResponse = await userInfoCall(
+          // Fetch from API using userListCall with current filters
+          const userDataResponse = await userListCall(
             accessToken,
-            null,
-            userRole,
-            true,
+            filters.user_id ? [filters.user_id] : null,
             currentPage,
-            defaultPageSize
+            defaultPageSize,
+            filters.email || null,
+            filters.user_role || null,
+            filters.team || null,
+            filters.sso_user_id || null,
+            filters.sort_by,
+            filters.sort_order
           );
 
           // Store in session storage
@@ -233,7 +374,7 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
           );
 
           setUserListResponse(userDataResponse);
-          setUserData(userDataResponse.users || []);
+          console.log("called from useEffect 2");
         }
 
         // Fetch roles if not cached
@@ -254,9 +395,9 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
       fetchData();
     }
 
-  }, [accessToken, token, userRole, userID, currentPage]);
+  }, [accessToken, token, userRole, userID]);
 
-  if (!userData) {
+  if (!userListResponse) {
     return <div>Loading...</div>;
   }
 
@@ -297,8 +438,164 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
           <TabPanel>
             <div className="bg-white rounded-lg shadow">
               <div className="border-b px-6 py-4">
-                <div className="flex flex-col md:flex-row items-start md:items-center justify-between space-y-4 md:space-y-0">
-                  <div className="flex items-center space-x-4">
+                <div className="flex flex-col space-y-4">
+                  {/* Search and Filter Controls */}
+                  <div className="flex flex-wrap items-center gap-3">
+                    {/* Email Search */}
+                    <div className="relative w-64">
+                      <input
+                        type="text"
+                        placeholder="Search by email..."
+                        className="w-full px-3 py-2 pl-8 border rounded-md text-sm focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-blue-500"
+                        value={filters.email}
+                        onChange={(e) => handleFilterChange('email', e.target.value)}
+                      />
+                      <svg
+                        className="absolute left-2.5 top-2.5 h-4 w-4 text-gray-500"
+                        fill="none"
+                        stroke="currentColor"
+                        viewBox="0 0 24 24"
+                      >
+                        <path
+                          strokeLinecap="round"
+                          strokeLinejoin="round"
+                          strokeWidth={2}
+                          d="M21 21l-6-6m2-5a7 7 0 11-14 0 7 7 0 0114 0z"
+                        />
+                      </svg>
+                    </div>
+
+                    {/* Filter Button */}
+                    <button
+                      className={`px-3 py-2 text-sm border rounded-md hover:bg-gray-50 flex items-center gap-2 ${showFilters ? 'bg-gray-100' : ''}`}
+                      onClick={() => setShowFilters(!showFilters)}
+                    >
+                      <svg
+                        className="w-4 h-4"
+                        fill="none"
+                        stroke="currentColor"
+                        viewBox="0 0 24 24"
+                      >
+                        <path
+                          strokeLinecap="round"
+                          strokeLinejoin="round"
+                          strokeWidth={2}
+                          d="M3 4a1 1 0 011-1h16a1 1 0 011 1v2.586a1 1 0 01-.293.707l-6.414 6.414a1 1 0 00-.293.707V17l-4 4v-6.586a1 1 0 00-.293-.707L3.293 7.293A1 1 0 013 6.586V4z"
+                        />
+                      </svg>
+                      Filters
+                      {(filters.user_id || filters.user_role || filters.team) && (
+                        <span className="w-2 h-2 rounded-full bg-blue-500"></span>
+                      )}
+                    </button>
+
+                    {/* Reset Filters Button */}
+                    <button
+                      className="px-3 py-2 text-sm border rounded-md hover:bg-gray-50 flex items-center gap-2"
+                      onClick={() => {
+                        setFilters({
+                          email: "",
+                          user_id: "",
+                          user_role: "",
+                          team: "",
+                          sso_user_id: "",
+                          model: "",
+                          min_spend: null,
+                          max_spend: null,
+                          sort_by: "created_at",
+                          sort_order: "desc"
+                        });
+                      }}
+                    >
+                      <svg
+                        className="w-4 h-4"
+                        fill="none"
+                        stroke="currentColor"
+                        viewBox="0 0 24 24"
+                      >
+                        <path
+                          strokeLinecap="round"
+                          strokeLinejoin="round"
+                          strokeWidth={2}
+                          d="M4 4v5h.582m15.356 2A8.001 8.001 0 004.582 9m0 0H9m11 11v-5h-.581m0 0a8.003 8.003 0 01-15.357-2m15.357 2H15"
+                        />
+                      </svg>
+                      Reset Filters
+                    </button>
+                  </div>
+
+                  {/* Additional Filters */}
+                  {showFilters && (
+                    <div className="flex flex-wrap items-center gap-3 mt-3">
+                      {/* User ID Search */}
+                      <div className="relative w-64">
+                        <input
+                          type="text"
+                          placeholder="Filter by User ID"
+                          className="w-full px-3 py-2 pl-8 border rounded-md text-sm focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-blue-500"
+                          value={filters.user_id}
+                          onChange={(e) => handleFilterChange('user_id', e.target.value)}
+                        />
+                        <svg
+                          className="absolute left-2.5 top-2.5 h-4 w-4 text-gray-500"
+                          fill="none"
+                          stroke="currentColor"
+                          viewBox="0 0 24 24"
+                        >
+                          <path
+                            strokeLinecap="round"
+                            strokeLinejoin="round"
+                            strokeWidth={2}
+                            d="M5.121 17.804A13.937 13.937 0 0112 16c2.5 0 4.847.655 6.879 1.804M15 10a3 3 0 11-6 0 3 3 0 016 0zm6 2a9 9 0 11-18 0 9 9 0 0118 0z"
+                          />
+                        </svg>
+                      </div>
+
+                      {/* Role Dropdown */}
+                      <div className="w-64">
+                        <Select
+                          value={filters.user_role}
+                          onValueChange={(value) => handleFilterChange('user_role', value)}
+                          placeholder="Select Role"
+                        >
+                          {Object.entries(possibleUIRoles).map(([key, value]) => (
+                            <SelectItem key={key} value={key}>
+                              {value.ui_label}
+                            </SelectItem>
+                          ))}
+                        </Select>
+                      </div>
+
+                      {/* Team Dropdown */}
+                      <div className="w-64">
+                        <Select
+                          value={filters.team}
+                          onValueChange={(value) => handleFilterChange('team', value)}
+                          placeholder="Select Team"
+                        >
+                          {teams?.map((team) => (
+                            <SelectItem key={team.team_id} value={team.team_id}>
+                              {team.team_alias || team.team_id}
+                            </SelectItem>
+                          ))}
+                        </Select>
+                      </div>
+                      
+                      {/* SSO ID Search */}
+                      <div className="relative w-64">
+                        <input
+                          type="text"
+                          placeholder="Filter by SSO ID"
+                          className="w-full px-3 py-2 pl-8 border rounded-md text-sm focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-blue-500"
+                          value={filters.sso_user_id}
+                          onChange={(e) => handleFilterChange('sso_user_id', e.target.value)}
+                        />
+                      </div>
+                    </div>
+                  )}
+
+                  {/* Results Count and Pagination */}
+                  <div className="flex justify-between items-center">
                     <span className="text-sm text-gray-700">
                       Showing{" "}
                       {userListResponse && userListResponse.users && userListResponse.users.length > 0
@@ -313,25 +610,28 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
                         : 0}{" "}
                       of {userListResponse ? userListResponse.total : 0} results
                     </span>
-                    <div className="flex items-center space-x-2">
+                    
+                    {/* Pagination Buttons */}
+                    <div className="flex space-x-2">
                       <button
-                        onClick={() => setCurrentPage((p) => Math.max(1, p - 1))}
-                        disabled={!userListResponse || currentPage <= 1}
-                        className="px-3 py-1 text-sm border rounded-md hover:bg-gray-50 disabled:opacity-50 disabled:cursor-not-allowed"
+                        onClick={() => handlePageChange(currentPage - 1)}
+                        disabled={currentPage === 1}
+                        className={`px-3 py-1 text-sm border rounded-md ${
+                          currentPage === 1
+                            ? 'bg-gray-100 text-gray-400 cursor-not-allowed'
+                            : 'hover:bg-gray-50'
+                        }`}
                       >
                         Previous
                       </button>
-                      <span className="text-sm text-gray-700">
-                        Page {userListResponse ? userListResponse.page : "-"} of{" "}
-                        {userListResponse ? userListResponse.total_pages : "-"}
-                      </span>
                       <button
-                        onClick={() => setCurrentPage((p) => p + 1)}
-                        disabled={
-                          !userListResponse ||
-                          currentPage >= userListResponse.total_pages
-                        }
-                        className="px-3 py-1 text-sm border rounded-md hover:bg-gray-50 disabled:opacity-50 disabled:cursor-not-allowed"
+                        onClick={() => handlePageChange(currentPage + 1)}
+                        disabled={!userListResponse || currentPage >= userListResponse.total_pages}
+                        className={`px-3 py-1 text-sm border rounded-md ${
+                          !userListResponse || currentPage >= userListResponse.total_pages
+                            ? 'bg-gray-100 text-gray-400 cursor-not-allowed'
+                            : 'hover:bg-gray-50'
+                        }`}
                       >
                         Next
                       </button>
@@ -339,10 +639,18 @@ const ViewUserDashboard: React.FC<ViewUserDashboardProps> = ({
                   </div>
                 </div>
               </div>
+
               <UserDataTable
-                data={userData || []}
+                data={userListResponse?.users || []}
                 columns={tableColumns}
-                isLoading={!userData}
+                isLoading={!userListResponse}
+                accessToken={accessToken}
+                userRole={userRole}
+                onSortChange={handleSortChange}
+                currentSort={{
+                  sortBy: filters.sort_by,
+                  sortOrder: filters.sort_order
+                }}
               />
             </div>
           </TabPanel>

ui/litellm-dashboard/src/components/view_users/table.tsx

@@ -18,21 +18,37 @@ import {
 } from "@tremor/react";
 import { SwitchVerticalIcon, ChevronUpIcon, ChevronDownIcon } from "@heroicons/react/outline";
 import { UserInfo } from "./types";
+import UserInfoView from "./user_info_view";
 
 interface UserDataTableProps {
   data: UserInfo[];
   columns: ColumnDef<UserInfo, any>[];
   isLoading?: boolean;
+  onSortChange?: (sortBy: string, sortOrder: 'asc' | 'desc') => void;
+  currentSort?: {
+    sortBy: string;
+    sortOrder: 'asc' | 'desc';
+  };
+  accessToken: string | null;
+  userRole: string | null;
 }
 
 export function UserDataTable({
   data = [],
   columns,
   isLoading = false,
+  onSortChange,
+  currentSort,
+  accessToken,
+  userRole,
 }: UserDataTableProps) {
   const [sorting, setSorting] = React.useState<SortingState>([
-    { id: "created_at", desc: true }
+    { 
+      id: currentSort?.sortBy || "created_at", 
+      desc: currentSort?.sortOrder === "desc" 
+    }
   ]);
+  const [selectedUserId, setSelectedUserId] = React.useState<string | null>(null);
 
   const table = useReactTable({
     data,
@@ -40,12 +56,49 @@ export function UserDataTable({
     state: {
       sorting,
     },
-    onSortingChange: setSorting,
+    onSortingChange: (newSorting: any) => {
+      setSorting(newSorting);
+      if (newSorting.length > 0) {
+        const sortState = newSorting[0];
+        const sortBy = sortState.id;
+        const sortOrder = sortState.desc ? 'desc' : 'asc';
+        onSortChange?.(sortBy, sortOrder);
+      }
+    },
     getCoreRowModel: getCoreRowModel(),
     getSortedRowModel: getSortedRowModel(),
     enableSorting: true,
   });
 
+  const handleUserClick = (userId: string) => {
+    setSelectedUserId(userId);
+  };
+
+  const handleCloseUserInfo = () => {
+    setSelectedUserId(null);
+  };
+
+  // Update local sorting state when currentSort prop changes
+  React.useEffect(() => {
+    if (currentSort) {
+      setSorting([{
+        id: currentSort.sortBy,
+        desc: currentSort.sortOrder === 'desc'
+      }]);
+    }
+  }, [currentSort]);
+
+  if (selectedUserId) {
+    return (
+      <UserInfoView
+        userId={selectedUserId}
+        onClose={handleCloseUserInfo}
+        accessToken={accessToken}
+        userRole={userRole}
+      />
+    );
+  }
+
   return (
     <div className="rounded-lg custom-border relative">
       <div className="overflow-x-auto">
@@ -110,6 +163,15 @@ export function UserDataTable({
                           ? 'sticky right-0 bg-white shadow-[-4px_0_8px_-6px_rgba(0,0,0,0.1)]'
                           : ''
                       }`}
+                      onClick={() => {
+                        if (cell.column.id === 'user_id') {
+                          handleUserClick(cell.getValue() as string);
+                        }
+                      }}
+                      style={{
+                        cursor: cell.column.id === 'user_id' ? 'pointer' : 'default',
+                        color: cell.column.id === 'user_id' ? '#3b82f6' : 'inherit',
+                      }}
                     >
                       {flexRender(cell.column.columnDef.cell, cell.getContext())}
                     </TableCell>

ui/litellm-dashboard/src/components/view_users/user_info_view.tsx

@@ -0,0 +1,297 @@
+import React, { useState } from "react";
+import {
+  Card,
+  Text,
+  Button,
+  Grid,
+  Col,
+  Tab,
+  TabList,
+  TabGroup,
+  TabPanel,
+  TabPanels,
+  Title,
+  Badge,
+} from "@tremor/react";
+import { ArrowLeftIcon, TrashIcon } from "@heroicons/react/outline";
+import { userInfoCall, userDeleteCall } from "../networking";
+import { message } from "antd";
+import { rolesWithWriteAccess } from '../../utils/roles';
+
+interface UserInfoViewProps {
+  userId: string;
+  onClose: () => void;
+  accessToken: string | null;
+  userRole: string | null;
+  onDelete?: () => void;
+}
+
+interface UserInfo {
+  user_id: string;
+  user_info: {
+    user_email: string | null;
+    user_role: string | null;
+    teams: any[] | null;
+    models: string[] | null;
+    max_budget: number | null;
+    spend: number | null;
+    metadata: Record<string, any> | null;
+    created_at: string | null;
+    updated_at: string | null;
+  };
+  keys: any[] | null;
+  teams: any[] | null;
+}
+
+export default function UserInfoView({ userId, onClose, accessToken, userRole, onDelete }: UserInfoViewProps) {
+  const [userData, setUserData] = useState<UserInfo | null>(null);
+  const [isDeleteModalOpen, setIsDeleteModalOpen] = useState(false);
+  const [isLoading, setIsLoading] = useState(true);
+
+  React.useEffect(() => {
+    console.log(`userId: ${userId}, userRole: ${userRole}, accessToken: ${accessToken}`)
+    const fetchUserData = async () => {
+      try {
+        if (!accessToken) return;
+        const data = await userInfoCall(accessToken, userId, userRole || "", false, null, null, true);
+        setUserData(data);
+      } catch (error) {
+        console.error("Error fetching user data:", error);
+        message.error("Failed to fetch user data");
+      } finally {
+        setIsLoading(false);
+      }
+    };
+
+    fetchUserData();
+  }, [accessToken, userId, userRole]);
+
+  const handleDelete = async () => {
+    try {
+      if (!accessToken) return;
+      await userDeleteCall(accessToken, [userId]);
+      message.success("User deleted successfully");
+      if (onDelete) {
+        onDelete();
+      }
+      onClose();
+    } catch (error) {
+      console.error("Error deleting user:", error);
+      message.error("Failed to delete user");
+    }
+  };
+
+  if (isLoading) {
+    return (
+      <div className="p-4">
+        <Button 
+          icon={ArrowLeftIcon} 
+          variant="light"
+          onClick={onClose}
+          className="mb-4"
+        >
+          Back to Users
+        </Button>
+        <Text>Loading user data...</Text>
+      </div>
+    );
+  }
+
+  if (!userData) {
+    return (
+      <div className="p-4">
+        <Button 
+          icon={ArrowLeftIcon} 
+          variant="light"
+          onClick={onClose}
+          className="mb-4"
+        >
+          Back to Users
+        </Button>
+        <Text>User not found</Text>
+      </div>
+    );
+  }
+
+  return (
+    <div className="p-4">
+      <div className="flex justify-between items-center mb-6">
+        <div>
+          <Button 
+            icon={ArrowLeftIcon} 
+            variant="light"
+            onClick={onClose}
+            className="mb-4"
+          >
+            Back to Users
+          </Button>
+          <Title>{userData.user_info?.user_email || "User"}</Title>
+          <Text className="text-gray-500 font-mono">{userData.user_id}</Text>
+        </div>
+        {userRole && rolesWithWriteAccess.includes(userRole) && (
+          <Button
+            icon={TrashIcon}
+            variant="secondary"
+            onClick={() => setIsDeleteModalOpen(true)}
+            className="flex items-center"
+          >
+            Delete User
+          </Button>
+        )}
+      </div>
+
+      {/* Delete Confirmation Modal */}
+      {isDeleteModalOpen && (
+        <div className="fixed z-10 inset-0 overflow-y-auto">
+          <div className="flex items-end justify-center min-h-screen pt-4 px-4 pb-20 text-center sm:block sm:p-0">
+            <div className="fixed inset-0 transition-opacity" aria-hidden="true">
+              <div className="absolute inset-0 bg-gray-500 opacity-75"></div>
+            </div>
+
+            <span className="hidden sm:inline-block sm:align-middle sm:h-screen" aria-hidden="true">&#8203;</span>
+
+            <div className="inline-block align-bottom bg-white rounded-lg text-left overflow-hidden shadow-xl transform transition-all sm:my-8 sm:align-middle sm:max-w-lg sm:w-full">
+              <div className="bg-white px-4 pt-5 pb-4 sm:p-6 sm:pb-4">
+                <div className="sm:flex sm:items-start">
+                  <div className="mt-3 text-center sm:mt-0 sm:ml-4 sm:text-left">
+                    <h3 className="text-lg leading-6 font-medium text-gray-900">
+                      Delete User
+                    </h3>
+                    <div className="mt-2">
+                      <p className="text-sm text-gray-500">
+                        Are you sure you want to delete this user?
+                      </p>
+                    </div>
+                  </div>
+                </div>
+              </div>
+              <div className="bg-gray-50 px-4 py-3 sm:px-6 sm:flex sm:flex-row-reverse">
+                <Button
+                  onClick={handleDelete}
+                  color="red"
+                  className="ml-2"
+                >
+                  Delete
+                </Button>
+                <Button onClick={() => setIsDeleteModalOpen(false)}>
+                  Cancel
+                </Button>
+              </div>
+            </div>
+          </div>
+        </div>
+      )}
+
+      <TabGroup>
+        <TabList className="mb-4">
+          <Tab>Overview</Tab>
+          <Tab>Details</Tab>
+        </TabList>
+
+        <TabPanels>
+          {/* Overview Panel */}
+          <TabPanel>
+            <Grid numItems={1} numItemsSm={2} numItemsLg={3} className="gap-6">
+              <Card>
+                <Text>Spend</Text>
+                <div className="mt-2">
+                  <Title>${Number(userData.user_info?.spend || 0).toFixed(4)}</Title>
+                  <Text>of {userData.user_info?.max_budget !== null ? `$${userData.user_info.max_budget}` : "Unlimited"}</Text>
+                </div>
+              </Card>
+
+              <Card>
+                <Text>Teams</Text>
+                <div className="mt-2">
+                  <Text>{userData.teams?.length || 0} teams</Text>
+                </div>
+              </Card>
+
+              <Card>
+                <Text>API Keys</Text>
+                <div className="mt-2">
+                  <Text>{userData.keys?.length || 0} keys</Text>
+                </div>
+              </Card>
+            </Grid>
+          </TabPanel>
+
+          {/* Details Panel */}
+          <TabPanel>
+            <Card>
+              <div className="space-y-4">
+                <div>
+                  <Text className="font-medium">User ID</Text>
+                  <Text className="font-mono">{userData.user_id}</Text>
+                </div>
+                
+                <div>
+                  <Text className="font-medium">Email</Text>
+                  <Text>{userData.user_info?.user_email || "Not Set"}</Text>
+                </div>
+
+                <div>
+                  <Text className="font-medium">Role</Text>
+                  <Text>{userData.user_info?.user_role || "Not Set"}</Text>
+                </div>
+
+                <div>
+                  <Text className="font-medium">Created</Text>
+                  <Text>{userData.user_info?.created_at ? new Date(userData.user_info.created_at).toLocaleString() : "Unknown"}</Text>
+                </div>
+
+                <div>
+                  <Text className="font-medium">Last Updated</Text>
+                  <Text>{userData.user_info?.updated_at ? new Date(userData.user_info.updated_at).toLocaleString() : "Unknown"}</Text>
+                </div>
+
+                <div>
+                  <Text className="font-medium">Teams</Text>
+                  <div className="flex flex-wrap gap-2 mt-1">
+                    {userData.teams?.length && userData.teams?.length > 0 ? (
+                      userData.teams?.map((team, index) => (
+                        <span
+                          key={index}
+                          className="px-2 py-1 bg-blue-100 rounded text-xs"
+                        >
+                          {team.team_alias || team.team_id}
+                        </span>
+                      ))
+                    ) : (
+                      <Text>No teams</Text>
+                    )}
+                  </div>
+                </div>
+
+                <div>
+                  <Text className="font-medium">API Keys</Text>
+                  <div className="flex flex-wrap gap-2 mt-1">
+                    {userData.keys?.length && userData.keys?.length > 0 ? (
+                      userData.keys.map((key, index) => (
+                        <span
+                          key={index}
+                          className="px-2 py-1 bg-green-100 rounded text-xs"
+                        >
+                          {key.key_alias || key.token}
+                        </span>
+                      ))
+                    ) : (
+                      <Text>No API keys</Text>
+                    )}
+                  </div>
+                </div>
+
+                <div>
+                  <Text className="font-medium">Metadata</Text>
+                  <pre className="bg-gray-100 p-2 rounded text-xs overflow-auto mt-1">
+                    {JSON.stringify(userData.user_info?.metadata || {}, null, 2)}
+                  </pre>
+                </div>
+              </div>
+            </Card>
+          </TabPanel>
+        </TabPanels>
+      </TabGroup>
+    </div>
+  );
+}