Merge remote-tracking branch 'upstream/main' into strip-telem

.github/actions/run-and-record-tests/action.yml

@@ -72,7 +72,8 @@ runs:
           echo "New recordings detected, committing and pushing"
           git add tests/integration/
 
-          git commit -m "Recordings update from CI (suite: ${{ inputs.suite }})"
+          git commit -m "Recordings update from CI (setup: ${{ inputs.setup }}, suite: ${{ inputs.suite }})"
+
           git fetch origin ${{ github.ref_name }}
           git rebase origin/${{ github.ref_name }}
           echo "Rebased successfully"
@@ -88,6 +89,8 @@ runs:
       run: |
         # Ollama logs (if ollama container exists)
         sudo docker logs ollama > ollama-${{ inputs.inference-mode }}.log 2>&1 || true
+        # vllm logs (if vllm container exists)
+        sudo docker logs vllm > vllm-${{ inputs.inference-mode }}.log 2>&1 || true
         # Note: distro container logs are now dumped in integration-tests.sh before container is removed
 
     - name: Upload logs

.github/actions/setup-vllm/action.yml

@@ -11,13 +11,14 @@ runs:
           --name vllm \
           -p 8000:8000 \
           --privileged=true \
-          quay.io/higginsd/vllm-cpu:65393ee064 \
+          quay.io/higginsd/vllm-cpu:65393ee064-qwen3 \
           --host 0.0.0.0 \
           --port 8000 \
           --enable-auto-tool-choice \
-          --tool-call-parser llama3_json \
+          --tool-call-parser hermes \
-          --model /root/.cache/Llama-3.2-1B-Instruct \
+          --model /root/.cache/Qwen3-0.6B \
-          --served-model-name meta-llama/Llama-3.2-1B-Instruct
+          --served-model-name Qwen/Qwen3-0.6B \
+          --max-model-len 8192
 
           # Wait for vllm to be ready
           echo "Waiting for vllm to be ready..."

.github/workflows/integration-tests.yml

@@ -27,7 +27,6 @@ on:
   schedule:
     # If changing the cron schedule, update the provider in the test-matrix job
     - cron: '0 0 * * *'  # (test latest client) Daily at 12 AM UTC
-    - cron: '1 0 * * 0'  # (test vllm) Weekly on Sunday at 1 AM UTC
   workflow_dispatch:
     inputs:
       test-all-client-versions:

client-sdks/stainless/openapi.yml

@@ -9976,6 +9976,70 @@ components:
         - metadata
       title: VectorStoreObject
       description: OpenAI Vector Store object.
+    VectorStoreChunkingStrategy:
+      oneOf:
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+      discriminator:
+        propertyName: type
+        mapping:
+          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+    VectorStoreChunkingStrategyAuto:
+      type: object
+      properties:
+        type:
+          type: string
+          const: auto
+          default: auto
+          description: >-
+            Strategy type, always "auto" for automatic chunking
+      additionalProperties: false
+      required:
+        - type
+      title: VectorStoreChunkingStrategyAuto
+      description: >-
+        Automatic chunking strategy for vector store files.
+    VectorStoreChunkingStrategyStatic:
+      type: object
+      properties:
+        type:
+          type: string
+          const: static
+          default: static
+          description: >-
+            Strategy type, always "static" for static chunking
+        static:
+          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
+          description: >-
+            Configuration parameters for the static chunking strategy
+      additionalProperties: false
+      required:
+        - type
+        - static
+      title: VectorStoreChunkingStrategyStatic
+      description: >-
+        Static chunking strategy with configurable parameters.
+    VectorStoreChunkingStrategyStaticConfig:
+      type: object
+      properties:
+        chunk_overlap_tokens:
+          type: integer
+          default: 400
+          description: >-
+            Number of tokens to overlap between adjacent chunks
+        max_chunk_size_tokens:
+          type: integer
+          default: 800
+          description: >-
+            Maximum number of tokens per chunk, must be between 100 and 4096
+      additionalProperties: false
+      required:
+        - chunk_overlap_tokens
+        - max_chunk_size_tokens
+      title: VectorStoreChunkingStrategyStaticConfig
+      description: >-
+        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreRequestWithExtraBody":
       type: object
       properties:
@@ -10001,15 +10065,7 @@ components:
           description: >-
             (Optional) Expiration policy for the vector store
         chunking_strategy:
-          type: object
+          $ref: '#/components/schemas/VectorStoreChunkingStrategy'
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
           description: >-
             (Optional) Strategy for splitting files into chunks
         metadata:
@@ -10085,70 +10141,6 @@ components:
         - deleted
       title: VectorStoreDeleteResponse
       description: Response from deleting a vector store.
-    VectorStoreChunkingStrategy:
-      oneOf:
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-      discriminator:
-        propertyName: type
-        mapping:
-          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-    VectorStoreChunkingStrategyAuto:
-      type: object
-      properties:
-        type:
-          type: string
-          const: auto
-          default: auto
-          description: >-
-            Strategy type, always "auto" for automatic chunking
-      additionalProperties: false
-      required:
-        - type
-      title: VectorStoreChunkingStrategyAuto
-      description: >-
-        Automatic chunking strategy for vector store files.
-    VectorStoreChunkingStrategyStatic:
-      type: object
-      properties:
-        type:
-          type: string
-          const: static
-          default: static
-          description: >-
-            Strategy type, always "static" for static chunking
-        static:
-          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
-          description: >-
-            Configuration parameters for the static chunking strategy
-      additionalProperties: false
-      required:
-        - type
-        - static
-      title: VectorStoreChunkingStrategyStatic
-      description: >-
-        Static chunking strategy with configurable parameters.
-    VectorStoreChunkingStrategyStaticConfig:
-      type: object
-      properties:
-        chunk_overlap_tokens:
-          type: integer
-          default: 400
-          description: >-
-            Number of tokens to overlap between adjacent chunks
-        max_chunk_size_tokens:
-          type: integer
-          default: 800
-          description: >-
-            Maximum number of tokens per chunk, must be between 100 and 4096
-      additionalProperties: false
-      required:
-        - chunk_overlap_tokens
-        - max_chunk_size_tokens
-      title: VectorStoreChunkingStrategyStaticConfig
-      description: >-
-        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreFileBatchRequestWithExtraBody":
       type: object
       properties:
@@ -10606,7 +10598,9 @@ components:
           description: >-
             Object type identifier for the search results page
         search_query:
-          type: string
+          type: array
+          items:
+            type: string
           description: >-
             The original search query that was executed
         data:

docs/docs/distributions/self_hosted_distro/starter.md

@@ -163,7 +163,41 @@ docker run \
   --port $LLAMA_STACK_PORT
 ```
 
-### Via venv
+The container will run the distribution with a SQLite store by default. This store is used for the following components:
+
+- Metadata store: store metadata about the models, providers, etc.
+- Inference store: collect of responses from the inference provider
+- Agents store: store agent configurations (sessions, turns, etc.)
+- Agents Responses store: store responses from the agents
+
+However, you can use PostgreSQL instead by running the `starter::run-with-postgres-store.yaml` configuration:
+
+```bash
+docker run \
+  -it \
+  --pull always \
+  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
+  -e OPENAI_API_KEY=your_openai_key \
+  -e FIREWORKS_API_KEY=your_fireworks_key \
+  -e TOGETHER_API_KEY=your_together_key \
+  -e POSTGRES_HOST=your_postgres_host \
+  -e POSTGRES_PORT=your_postgres_port \
+  -e POSTGRES_DB=your_postgres_db \
+  -e POSTGRES_USER=your_postgres_user \
+  -e POSTGRES_PASSWORD=your_postgres_password \
+  llamastack/distribution-starter \
+  starter::run-with-postgres-store.yaml
+```
+
+Postgres environment variables:
+
+- `POSTGRES_HOST`: Postgres host (default: `localhost`)
+- `POSTGRES_PORT`: Postgres port (default: `5432`)
+- `POSTGRES_DB`: Postgres database name (default: `llamastack`)
+- `POSTGRES_USER`: Postgres username (default: `llamastack`)
+- `POSTGRES_PASSWORD`: Postgres password (default: `llamastack`)
+
+### Via Conda or venv
 
 Ensure you have configured the starter distribution using the environment variables explained above.
 
@@ -171,8 +205,11 @@ Ensure you have configured the starter distribution using the environment variab
 # Install dependencies for the starter distribution
 uv run --with llama-stack llama stack list-deps starter | xargs -L1 uv pip install
 
-# Run the server
+# Run the server (with SQLite - default)
 uv run --with llama-stack llama stack run starter
+
+# Or run with PostgreSQL
+uv run --with llama-stack llama stack run starter::run-with-postgres-store.yaml
 ```
 
 ## Example Usage

docs/docs/providers/inference/remote_passthrough.mdx

@@ -16,7 +16,7 @@ Passthrough inference provider for connecting to any external inference service
 |-------|------|----------|---------|-------------|
 | `allowed_models` | `list[str \| None` | No |  | List of models that should be registered with the model registry. If None, all models are allowed. |
 | `refresh_models` | `<class 'bool'>` | No | False | Whether to refresh models periodically from the provider |
-| `api_key` | `pydantic.types.SecretStr \| None` | No |  | API Key for the passthrouth endpoint |
+| `api_key` | `pydantic.types.SecretStr \| None` | No |  | Authentication credential for the provider |
 | `url` | `<class 'str'>` | No |  | The URL for the passthrough endpoint |
 
 ## Sample Configuration

docs/static/llama-stack-spec.yaml

@@ -9260,6 +9260,70 @@ components:
         - metadata
       title: VectorStoreObject
       description: OpenAI Vector Store object.
+    VectorStoreChunkingStrategy:
+      oneOf:
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+      discriminator:
+        propertyName: type
+        mapping:
+          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+    VectorStoreChunkingStrategyAuto:
+      type: object
+      properties:
+        type:
+          type: string
+          const: auto
+          default: auto
+          description: >-
+            Strategy type, always "auto" for automatic chunking
+      additionalProperties: false
+      required:
+        - type
+      title: VectorStoreChunkingStrategyAuto
+      description: >-
+        Automatic chunking strategy for vector store files.
+    VectorStoreChunkingStrategyStatic:
+      type: object
+      properties:
+        type:
+          type: string
+          const: static
+          default: static
+          description: >-
+            Strategy type, always "static" for static chunking
+        static:
+          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
+          description: >-
+            Configuration parameters for the static chunking strategy
+      additionalProperties: false
+      required:
+        - type
+        - static
+      title: VectorStoreChunkingStrategyStatic
+      description: >-
+        Static chunking strategy with configurable parameters.
+    VectorStoreChunkingStrategyStaticConfig:
+      type: object
+      properties:
+        chunk_overlap_tokens:
+          type: integer
+          default: 400
+          description: >-
+            Number of tokens to overlap between adjacent chunks
+        max_chunk_size_tokens:
+          type: integer
+          default: 800
+          description: >-
+            Maximum number of tokens per chunk, must be between 100 and 4096
+      additionalProperties: false
+      required:
+        - chunk_overlap_tokens
+        - max_chunk_size_tokens
+      title: VectorStoreChunkingStrategyStaticConfig
+      description: >-
+        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreRequestWithExtraBody":
       type: object
       properties:
@@ -9285,15 +9349,7 @@ components:
           description: >-
             (Optional) Expiration policy for the vector store
         chunking_strategy:
-          type: object
+          $ref: '#/components/schemas/VectorStoreChunkingStrategy'
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
           description: >-
             (Optional) Strategy for splitting files into chunks
         metadata:
@@ -9369,70 +9425,6 @@ components:
         - deleted
       title: VectorStoreDeleteResponse
       description: Response from deleting a vector store.
-    VectorStoreChunkingStrategy:
-      oneOf:
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-      discriminator:
-        propertyName: type
-        mapping:
-          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-    VectorStoreChunkingStrategyAuto:
-      type: object
-      properties:
-        type:
-          type: string
-          const: auto
-          default: auto
-          description: >-
-            Strategy type, always "auto" for automatic chunking
-      additionalProperties: false
-      required:
-        - type
-      title: VectorStoreChunkingStrategyAuto
-      description: >-
-        Automatic chunking strategy for vector store files.
-    VectorStoreChunkingStrategyStatic:
-      type: object
-      properties:
-        type:
-          type: string
-          const: static
-          default: static
-          description: >-
-            Strategy type, always "static" for static chunking
-        static:
-          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
-          description: >-
-            Configuration parameters for the static chunking strategy
-      additionalProperties: false
-      required:
-        - type
-        - static
-      title: VectorStoreChunkingStrategyStatic
-      description: >-
-        Static chunking strategy with configurable parameters.
-    VectorStoreChunkingStrategyStaticConfig:
-      type: object
-      properties:
-        chunk_overlap_tokens:
-          type: integer
-          default: 400
-          description: >-
-            Number of tokens to overlap between adjacent chunks
-        max_chunk_size_tokens:
-          type: integer
-          default: 800
-          description: >-
-            Maximum number of tokens per chunk, must be between 100 and 4096
-      additionalProperties: false
-      required:
-        - chunk_overlap_tokens
-        - max_chunk_size_tokens
-      title: VectorStoreChunkingStrategyStaticConfig
-      description: >-
-        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreFileBatchRequestWithExtraBody":
       type: object
       properties:
@@ -9890,7 +9882,9 @@ components:
           description: >-
             Object type identifier for the search results page
         search_query:
-          type: string
+          type: array
+          items:
+            type: string
           description: >-
             The original search query that was executed
         data:

docs/static/stainless-llama-stack-spec.yaml

@@ -9976,6 +9976,70 @@ components:
         - metadata
       title: VectorStoreObject
       description: OpenAI Vector Store object.
+    VectorStoreChunkingStrategy:
+      oneOf:
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+      discriminator:
+        propertyName: type
+        mapping:
+          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
+          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
+    VectorStoreChunkingStrategyAuto:
+      type: object
+      properties:
+        type:
+          type: string
+          const: auto
+          default: auto
+          description: >-
+            Strategy type, always "auto" for automatic chunking
+      additionalProperties: false
+      required:
+        - type
+      title: VectorStoreChunkingStrategyAuto
+      description: >-
+        Automatic chunking strategy for vector store files.
+    VectorStoreChunkingStrategyStatic:
+      type: object
+      properties:
+        type:
+          type: string
+          const: static
+          default: static
+          description: >-
+            Strategy type, always "static" for static chunking
+        static:
+          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
+          description: >-
+            Configuration parameters for the static chunking strategy
+      additionalProperties: false
+      required:
+        - type
+        - static
+      title: VectorStoreChunkingStrategyStatic
+      description: >-
+        Static chunking strategy with configurable parameters.
+    VectorStoreChunkingStrategyStaticConfig:
+      type: object
+      properties:
+        chunk_overlap_tokens:
+          type: integer
+          default: 400
+          description: >-
+            Number of tokens to overlap between adjacent chunks
+        max_chunk_size_tokens:
+          type: integer
+          default: 800
+          description: >-
+            Maximum number of tokens per chunk, must be between 100 and 4096
+      additionalProperties: false
+      required:
+        - chunk_overlap_tokens
+        - max_chunk_size_tokens
+      title: VectorStoreChunkingStrategyStaticConfig
+      description: >-
+        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreRequestWithExtraBody":
       type: object
       properties:
@@ -10001,15 +10065,7 @@ components:
           description: >-
             (Optional) Expiration policy for the vector store
         chunking_strategy:
-          type: object
+          $ref: '#/components/schemas/VectorStoreChunkingStrategy'
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
           description: >-
             (Optional) Strategy for splitting files into chunks
         metadata:
@@ -10085,70 +10141,6 @@ components:
         - deleted
       title: VectorStoreDeleteResponse
       description: Response from deleting a vector store.
-    VectorStoreChunkingStrategy:
-      oneOf:
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-        - $ref: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-      discriminator:
-        propertyName: type
-        mapping:
-          auto: '#/components/schemas/VectorStoreChunkingStrategyAuto'
-          static: '#/components/schemas/VectorStoreChunkingStrategyStatic'
-    VectorStoreChunkingStrategyAuto:
-      type: object
-      properties:
-        type:
-          type: string
-          const: auto
-          default: auto
-          description: >-
-            Strategy type, always "auto" for automatic chunking
-      additionalProperties: false
-      required:
-        - type
-      title: VectorStoreChunkingStrategyAuto
-      description: >-
-        Automatic chunking strategy for vector store files.
-    VectorStoreChunkingStrategyStatic:
-      type: object
-      properties:
-        type:
-          type: string
-          const: static
-          default: static
-          description: >-
-            Strategy type, always "static" for static chunking
-        static:
-          $ref: '#/components/schemas/VectorStoreChunkingStrategyStaticConfig'
-          description: >-
-            Configuration parameters for the static chunking strategy
-      additionalProperties: false
-      required:
-        - type
-        - static
-      title: VectorStoreChunkingStrategyStatic
-      description: >-
-        Static chunking strategy with configurable parameters.
-    VectorStoreChunkingStrategyStaticConfig:
-      type: object
-      properties:
-        chunk_overlap_tokens:
-          type: integer
-          default: 400
-          description: >-
-            Number of tokens to overlap between adjacent chunks
-        max_chunk_size_tokens:
-          type: integer
-          default: 800
-          description: >-
-            Maximum number of tokens per chunk, must be between 100 and 4096
-      additionalProperties: false
-      required:
-        - chunk_overlap_tokens
-        - max_chunk_size_tokens
-      title: VectorStoreChunkingStrategyStaticConfig
-      description: >-
-        Configuration for static chunking strategy.
     "OpenAICreateVectorStoreFileBatchRequestWithExtraBody":
       type: object
       properties:
@@ -10606,7 +10598,9 @@ components:
           description: >-
             Object type identifier for the search results page
         search_query:
-          type: string
+          type: array
+          items:
+            type: string
           description: >-
             The original search query that was executed
         data:

scripts/integration-tests.sh

@@ -405,11 +405,6 @@ fi
 echo "=== Running Integration Tests ==="
 EXCLUDE_TESTS="builtin_tool or safety_with_image or code_interpreter or test_rag"
 
-# Additional exclusions for vllm setup
-if [[ "$TEST_SETUP" == "vllm" ]]; then
-    EXCLUDE_TESTS="${EXCLUDE_TESTS} or test_inference_store_tool_calls"
-fi
-
 PYTEST_PATTERN="not( $EXCLUDE_TESTS )"
 if [[ -n "$TEST_PATTERN" ]]; then
     PYTEST_PATTERN="${PYTEST_PATTERN} and $TEST_PATTERN"

src/llama_stack/apis/vector_io/vector_io.py

@@ -260,7 +260,7 @@ class VectorStoreSearchResponsePage(BaseModel):
     """
 
     object: str = "vector_store.search_results.page"
-    search_query: str
+    search_query: list[str]
     data: list[VectorStoreSearchResponse]
     has_more: bool = False
     next_page: str | None = None
@@ -478,7 +478,7 @@ class OpenAICreateVectorStoreRequestWithExtraBody(BaseModel, extra="allow"):
     name: str | None = None
     file_ids: list[str] | None = None
     expires_after: dict[str, Any] | None = None
-    chunking_strategy: dict[str, Any] | None = None
+    chunking_strategy: VectorStoreChunkingStrategy | None = None
     metadata: dict[str, Any] | None = None
 
 

src/llama_stack/core/routers/vector_io.py

@@ -20,6 +20,8 @@ from llama_stack.apis.vector_io import (
     SearchRankingOptions,
     VectorIO,
     VectorStoreChunkingStrategy,
+    VectorStoreChunkingStrategyStatic,
+    VectorStoreChunkingStrategyStaticConfig,
     VectorStoreDeleteResponse,
     VectorStoreFileBatchObject,
     VectorStoreFileContentsResponse,
@@ -167,6 +169,13 @@ class VectorIORouter(VectorIO):
         if embedding_dimension is not None:
             params.model_extra["embedding_dimension"] = embedding_dimension
 
+        # Set chunking strategy explicitly if not provided
+        if params.chunking_strategy is None or params.chunking_strategy.type == "auto":
+            # actualize the chunking strategy to static
+            params.chunking_strategy = VectorStoreChunkingStrategyStatic(
+                static=VectorStoreChunkingStrategyStaticConfig()
+            )
+
         return await provider.openai_create_vector_store(params)
 
     async def openai_list_vector_stores(
@@ -283,6 +292,8 @@ class VectorIORouter(VectorIO):
         chunking_strategy: VectorStoreChunkingStrategy | None = None,
     ) -> VectorStoreFileObject:
         logger.debug(f"VectorIORouter.openai_attach_file_to_vector_store: {vector_store_id}, {file_id}")
+        if chunking_strategy is None or chunking_strategy.type == "auto":
+            chunking_strategy = VectorStoreChunkingStrategyStatic(static=VectorStoreChunkingStrategyStaticConfig())
         provider = await self.routing_table.get_provider_impl(vector_store_id)
         return await provider.openai_attach_file_to_vector_store(
             vector_store_id=vector_store_id,

src/llama_stack/core/utils/config_resolution.py

@@ -52,7 +52,17 @@ def resolve_config_or_distro(
             logger.debug(f"Using distribution: {distro_config}")
             return distro_config
 
-    # Strategy 3: Try as built distribution name
+    # Strategy 3: Try as distro config path (if no .yaml extension and contains a slash)
+    # eg: starter::run-with-postgres-store.yaml
+    # Use :: to avoid slash and confusion with a filesystem path
+    if "::" in config_or_distro:
+        distro_name, config_name = config_or_distro.split("::")
+        distro_config = _get_distro_config_path(distro_name, config_name)
+        if distro_config.exists():
+            logger.info(f"Using distribution: {distro_config}")
+            return distro_config
+
+    # Strategy 4: Try as built distribution name
     distrib_config = DISTRIBS_BASE_DIR / f"llamastack-{config_or_distro}" / f"{config_or_distro}-{mode}.yaml"
     if distrib_config.exists():
         logger.debug(f"Using built distribution: {distrib_config}")
@@ -63,13 +73,15 @@ def resolve_config_or_distro(
         logger.debug(f"Using built distribution: {distrib_config}")
         return distrib_config
 
-    # Strategy 4: Failed - provide helpful error
+    # Strategy 5: Failed - provide helpful error
     raise ValueError(_format_resolution_error(config_or_distro, mode))
 
 
-def _get_distro_config_path(distro_name: str, mode: Mode) -> Path:
+def _get_distro_config_path(distro_name: str, mode: str) -> Path:
     """Get the config file path for a distro."""
-    return DISTRO_DIR / distro_name / f"{mode}.yaml"
+    if not mode.endswith(".yaml"):
+        mode = f"{mode}.yaml"
+    return DISTRO_DIR / distro_name / mode
 
 
 def _format_resolution_error(config_or_distro: str, mode: Mode) -> str:

src/llama_stack/core/utils/exec.py

@@ -84,6 +84,15 @@ def run_command(command: list[str]) -> int:
             text=True,
             check=False,
         )
+
+        # Print stdout and stderr if command failed
+        if result.returncode != 0:
+            log.error(f"Command {' '.join(command)} failed with returncode {result.returncode}")
+            if result.stdout:
+                log.error(f"STDOUT: {result.stdout}")
+            if result.stderr:
+                log.error(f"STDERR: {result.stderr}")
+
         return result.returncode
     except subprocess.SubprocessError as e:
         log.error(f"Subprocess error: {e}")

src/llama_stack/distributions/ci-tests/build.yaml

@@ -56,4 +56,5 @@ image_type: venv
 additional_pip_packages:
 - aiosqlite
 - asyncpg
+- psycopg2-binary
 - sqlalchemy[asyncio]

src/llama_stack/distributions/ci-tests/ci_tests.py

@@ -13,5 +13,6 @@ from ..starter.starter import get_distribution_template as get_starter_distribut
 def get_distribution_template() -> DistributionTemplate:
     template = get_starter_distribution_template(name="ci-tests")
     template.description = "CI tests for Llama Stack"
+    template.run_configs.pop("run-with-postgres-store.yaml", None)
 
     return template

src/llama_stack/distributions/postgres-demo/__init__.py

@@ -1,7 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-from .postgres_demo import get_distribution_template  # noqa: F401

src/llama_stack/distributions/postgres-demo/build.yaml

@@ -1,23 +0,0 @@
-version: 2
-distribution_spec:
-  description: Quick start template for running Llama Stack with several popular providers
-  providers:
-    inference:
-    - provider_type: remote::vllm
-    - provider_type: inline::sentence-transformers
-    vector_io:
-    - provider_type: remote::chromadb
-    safety:
-    - provider_type: inline::llama-guard
-    agents:
-    - provider_type: inline::meta-reference
-    tool_runtime:
-    - provider_type: remote::brave-search
-    - provider_type: remote::tavily-search
-    - provider_type: inline::rag-runtime
-    - provider_type: remote::model-context-protocol
-image_type: venv
-additional_pip_packages:
-- asyncpg
-- psycopg2-binary
-- sqlalchemy[asyncio]

src/llama_stack/distributions/postgres-demo/postgres_demo.py

@@ -1,125 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-
-from llama_stack.apis.models import ModelType
-from llama_stack.core.datatypes import (
-    BuildProvider,
-    ModelInput,
-    Provider,
-    ShieldInput,
-    ToolGroupInput,
-)
-from llama_stack.distributions.template import (
-    DistributionTemplate,
-    RunConfigSettings,
-)
-from llama_stack.providers.inline.inference.sentence_transformers import SentenceTransformersInferenceConfig
-from llama_stack.providers.remote.inference.vllm import VLLMInferenceAdapterConfig
-from llama_stack.providers.remote.vector_io.chroma.config import ChromaVectorIOConfig
-from llama_stack.providers.utils.kvstore.config import PostgresKVStoreConfig
-from llama_stack.providers.utils.sqlstore.sqlstore import PostgresSqlStoreConfig
-
-
-def get_distribution_template() -> DistributionTemplate:
-    inference_providers = [
-        Provider(
-            provider_id="vllm-inference",
-            provider_type="remote::vllm",
-            config=VLLMInferenceAdapterConfig.sample_run_config(
-                url="${env.VLLM_URL:=http://localhost:8000/v1}",
-            ),
-        ),
-    ]
-    providers = {
-        "inference": [
-            BuildProvider(provider_type="remote::vllm"),
-            BuildProvider(provider_type="inline::sentence-transformers"),
-        ],
-        "vector_io": [BuildProvider(provider_type="remote::chromadb")],
-        "safety": [BuildProvider(provider_type="inline::llama-guard")],
-        "agents": [BuildProvider(provider_type="inline::meta-reference")],
-        "tool_runtime": [
-            BuildProvider(provider_type="remote::brave-search"),
-            BuildProvider(provider_type="remote::tavily-search"),
-            BuildProvider(provider_type="inline::rag-runtime"),
-            BuildProvider(provider_type="remote::model-context-protocol"),
-        ],
-    }
-    name = "postgres-demo"
-
-    vector_io_providers = [
-        Provider(
-            provider_id="${env.ENABLE_CHROMADB:+chromadb}",
-            provider_type="remote::chromadb",
-            config=ChromaVectorIOConfig.sample_run_config(
-                f"~/.llama/distributions/{name}",
-                url="${env.CHROMADB_URL:=}",
-            ),
-        ),
-    ]
-    default_tool_groups = [
-        ToolGroupInput(
-            toolgroup_id="builtin::websearch",
-            provider_id="tavily-search",
-        ),
-        ToolGroupInput(
-            toolgroup_id="builtin::rag",
-            provider_id="rag-runtime",
-        ),
-    ]
-
-    default_models = [
-        ModelInput(
-            model_id="${env.INFERENCE_MODEL}",
-            provider_id="vllm-inference",
-        )
-    ]
-    embedding_provider = Provider(
-        provider_id="sentence-transformers",
-        provider_type="inline::sentence-transformers",
-        config=SentenceTransformersInferenceConfig.sample_run_config(),
-    )
-    embedding_model = ModelInput(
-        model_id="nomic-embed-text-v1.5",
-        provider_id=embedding_provider.provider_id,
-        model_type=ModelType.embedding,
-        metadata={
-            "embedding_dimension": 768,
-        },
-    )
-    return DistributionTemplate(
-        name=name,
-        distro_type="self_hosted",
-        description="Quick start template for running Llama Stack with several popular providers",
-        container_image=None,
-        template_path=None,
-        providers=providers,
-        available_models_by_provider={},
-        run_configs={
-            "run.yaml": RunConfigSettings(
-                provider_overrides={
-                    "inference": inference_providers + [embedding_provider],
-                    "vector_io": vector_io_providers,
-                },
-                default_models=default_models + [embedding_model],
-                default_tool_groups=default_tool_groups,
-                default_shields=[ShieldInput(shield_id="meta-llama/Llama-Guard-3-8B")],
-                storage_backends={
-                    "kv_default": PostgresKVStoreConfig.sample_run_config(
-                        table_name="llamastack_kvstore",
-                    ),
-                    "sql_default": PostgresSqlStoreConfig.sample_run_config(),
-                },
-            ),
-        },
-        run_config_env_vars={
-            "LLAMA_STACK_PORT": (
-                "8321",
-                "Port for the Llama Stack distribution server",
-            ),
-        },
-    )

src/llama_stack/distributions/starter-gpu/build.yaml

@@ -57,4 +57,5 @@ image_type: venv
 additional_pip_packages:
 - aiosqlite
 - asyncpg
+- psycopg2-binary
 - sqlalchemy[asyncio]

src/llama_stack/distributions/starter-gpu/run-with-postgres-store.yaml

@@ -0,0 +1,281 @@
+version: 2
+image_name: starter-gpu
+apis:
+- agents
+- batches
+- datasetio
+- eval
+- files
+- inference
+- post_training
+- safety
+- scoring
+- tool_runtime
+- vector_io
+providers:
+  inference:
+  - provider_id: ${env.CEREBRAS_API_KEY:+cerebras}
+    provider_type: remote::cerebras
+    config:
+      base_url: https://api.cerebras.ai
+      api_key: ${env.CEREBRAS_API_KEY:=}
+  - provider_id: ${env.OLLAMA_URL:+ollama}
+    provider_type: remote::ollama
+    config:
+      url: ${env.OLLAMA_URL:=http://localhost:11434}
+  - provider_id: ${env.VLLM_URL:+vllm}
+    provider_type: remote::vllm
+    config:
+      url: ${env.VLLM_URL:=}
+      max_tokens: ${env.VLLM_MAX_TOKENS:=4096}
+      api_token: ${env.VLLM_API_TOKEN:=fake}
+      tls_verify: ${env.VLLM_TLS_VERIFY:=true}
+  - provider_id: ${env.TGI_URL:+tgi}
+    provider_type: remote::tgi
+    config:
+      url: ${env.TGI_URL:=}
+  - provider_id: fireworks
+    provider_type: remote::fireworks
+    config:
+      url: https://api.fireworks.ai/inference/v1
+      api_key: ${env.FIREWORKS_API_KEY:=}
+  - provider_id: together
+    provider_type: remote::together
+    config:
+      url: https://api.together.xyz/v1
+      api_key: ${env.TOGETHER_API_KEY:=}
+  - provider_id: bedrock
+    provider_type: remote::bedrock
+  - provider_id: ${env.NVIDIA_API_KEY:+nvidia}
+    provider_type: remote::nvidia
+    config:
+      url: ${env.NVIDIA_BASE_URL:=https://integrate.api.nvidia.com}
+      api_key: ${env.NVIDIA_API_KEY:=}
+      append_api_version: ${env.NVIDIA_APPEND_API_VERSION:=True}
+  - provider_id: openai
+    provider_type: remote::openai
+    config:
+      api_key: ${env.OPENAI_API_KEY:=}
+      base_url: ${env.OPENAI_BASE_URL:=https://api.openai.com/v1}
+  - provider_id: anthropic
+    provider_type: remote::anthropic
+    config:
+      api_key: ${env.ANTHROPIC_API_KEY:=}
+  - provider_id: gemini
+    provider_type: remote::gemini
+    config:
+      api_key: ${env.GEMINI_API_KEY:=}
+  - provider_id: ${env.VERTEX_AI_PROJECT:+vertexai}
+    provider_type: remote::vertexai
+    config:
+      project: ${env.VERTEX_AI_PROJECT:=}
+      location: ${env.VERTEX_AI_LOCATION:=us-central1}
+  - provider_id: groq
+    provider_type: remote::groq
+    config:
+      url: https://api.groq.com
+      api_key: ${env.GROQ_API_KEY:=}
+  - provider_id: sambanova
+    provider_type: remote::sambanova
+    config:
+      url: https://api.sambanova.ai/v1
+      api_key: ${env.SAMBANOVA_API_KEY:=}
+  - provider_id: ${env.AZURE_API_KEY:+azure}
+    provider_type: remote::azure
+    config:
+      api_key: ${env.AZURE_API_KEY:=}
+      api_base: ${env.AZURE_API_BASE:=}
+      api_version: ${env.AZURE_API_VERSION:=}
+      api_type: ${env.AZURE_API_TYPE:=}
+  - provider_id: sentence-transformers
+    provider_type: inline::sentence-transformers
+  vector_io:
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      persistence:
+        namespace: vector_io::faiss
+        backend: kv_default
+  - provider_id: sqlite-vec
+    provider_type: inline::sqlite-vec
+    config:
+      db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/starter-gpu}/sqlite_vec.db
+      persistence:
+        namespace: vector_io::sqlite_vec
+        backend: kv_default
+  - provider_id: ${env.MILVUS_URL:+milvus}
+    provider_type: inline::milvus
+    config:
+      db_path: ${env.MILVUS_DB_PATH:=~/.llama/distributions/starter-gpu}/milvus.db
+      persistence:
+        namespace: vector_io::milvus
+        backend: kv_default
+  - provider_id: ${env.CHROMADB_URL:+chromadb}
+    provider_type: remote::chromadb
+    config:
+      url: ${env.CHROMADB_URL:=}
+      persistence:
+        namespace: vector_io::chroma_remote
+        backend: kv_default
+  - provider_id: ${env.PGVECTOR_DB:+pgvector}
+    provider_type: remote::pgvector
+    config:
+      host: ${env.PGVECTOR_HOST:=localhost}
+      port: ${env.PGVECTOR_PORT:=5432}
+      db: ${env.PGVECTOR_DB:=}
+      user: ${env.PGVECTOR_USER:=}
+      password: ${env.PGVECTOR_PASSWORD:=}
+      persistence:
+        namespace: vector_io::pgvector
+        backend: kv_default
+  - provider_id: ${env.QDRANT_URL:+qdrant}
+    provider_type: remote::qdrant
+    config:
+      api_key: ${env.QDRANT_API_KEY:=}
+      persistence:
+        namespace: vector_io::qdrant_remote
+        backend: kv_default
+  - provider_id: ${env.WEAVIATE_CLUSTER_URL:+weaviate}
+    provider_type: remote::weaviate
+    config:
+      weaviate_api_key: null
+      weaviate_cluster_url: ${env.WEAVIATE_CLUSTER_URL:=localhost:8080}
+      persistence:
+        namespace: vector_io::weaviate
+        backend: kv_default
+  files:
+  - provider_id: meta-reference-files
+    provider_type: inline::localfs
+    config:
+      storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/distributions/starter-gpu/files}
+      metadata_store:
+        table_name: files_metadata
+        backend: sql_default
+  safety:
+  - provider_id: llama-guard
+    provider_type: inline::llama-guard
+    config:
+      excluded_categories: []
+  - provider_id: code-scanner
+    provider_type: inline::code-scanner
+  agents:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      persistence_store:
+        type: sql_postgres
+        host: ${env.POSTGRES_HOST:=localhost}
+        port: ${env.POSTGRES_PORT:=5432}
+        db: ${env.POSTGRES_DB:=llamastack}
+        user: ${env.POSTGRES_USER:=llamastack}
+        password: ${env.POSTGRES_PASSWORD:=llamastack}
+      responses_store:
+        type: sql_postgres
+        host: ${env.POSTGRES_HOST:=localhost}
+        port: ${env.POSTGRES_PORT:=5432}
+        db: ${env.POSTGRES_DB:=llamastack}
+        user: ${env.POSTGRES_USER:=llamastack}
+        password: ${env.POSTGRES_PASSWORD:=llamastack}
+  post_training:
+  - provider_id: huggingface-gpu
+    provider_type: inline::huggingface-gpu
+    config:
+      checkpoint_format: huggingface
+      distributed_backend: null
+      device: cpu
+      dpo_output_dir: ~/.llama/distributions/starter-gpu/dpo_output
+  eval:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      kvstore:
+        namespace: eval
+        backend: kv_default
+  datasetio:
+  - provider_id: huggingface
+    provider_type: remote::huggingface
+    config:
+      kvstore:
+        namespace: datasetio::huggingface
+        backend: kv_default
+  - provider_id: localfs
+    provider_type: inline::localfs
+    config:
+      kvstore:
+        namespace: datasetio::localfs
+        backend: kv_default
+  scoring:
+  - provider_id: basic
+    provider_type: inline::basic
+  - provider_id: llm-as-judge
+    provider_type: inline::llm-as-judge
+  - provider_id: braintrust
+    provider_type: inline::braintrust
+    config:
+      openai_api_key: ${env.OPENAI_API_KEY:=}
+  tool_runtime:
+  - provider_id: brave-search
+    provider_type: remote::brave-search
+    config:
+      api_key: ${env.BRAVE_SEARCH_API_KEY:=}
+      max_results: 3
+  - provider_id: tavily-search
+    provider_type: remote::tavily-search
+    config:
+      api_key: ${env.TAVILY_SEARCH_API_KEY:=}
+      max_results: 3
+  - provider_id: rag-runtime
+    provider_type: inline::rag-runtime
+  - provider_id: model-context-protocol
+    provider_type: remote::model-context-protocol
+  batches:
+  - provider_id: reference
+    provider_type: inline::reference
+    config:
+      kvstore:
+        namespace: batches
+        backend: kv_postgres
+storage:
+  backends:
+    kv_postgres:
+      type: kv_postgres
+      host: ${env.POSTGRES_HOST:=localhost}
+      port: ${env.POSTGRES_PORT:=5432}
+      db: ${env.POSTGRES_DB:=llamastack}
+      user: ${env.POSTGRES_USER:=llamastack}
+      password: ${env.POSTGRES_PASSWORD:=llamastack}
+      table_name: ${env.POSTGRES_TABLE_NAME:=llamastack_kvstore}
+    sql_postgres:
+      type: sql_postgres
+      host: ${env.POSTGRES_HOST:=localhost}
+      port: ${env.POSTGRES_PORT:=5432}
+      db: ${env.POSTGRES_DB:=llamastack}
+      user: ${env.POSTGRES_USER:=llamastack}
+      password: ${env.POSTGRES_PASSWORD:=llamastack}
+  stores:
+    metadata:
+      namespace: registry
+      backend: kv_postgres
+    inference:
+      table_name: inference_store
+      backend: sql_postgres
+      max_write_queue_size: 10000
+      num_writers: 4
+    conversations:
+      table_name: openai_conversations
+      backend: sql_postgres
+    prompts:
+      namespace: prompts
+      backend: kv_postgres
+registered_resources:
+  models: []
+  shields: []
+  vector_dbs: []
+  datasets: []
+  scoring_fns: []
+  benchmarks: []
+  tool_groups: []
+server:
+  port: 8321
+telemetry:
+  enabled: true

src/llama_stack/distributions/starter/build.yaml

@@ -57,4 +57,5 @@ image_type: venv
 additional_pip_packages:
 - aiosqlite
 - asyncpg
+- psycopg2-binary
 - sqlalchemy[asyncio]

src/llama_stack/distributions/starter/run-with-postgres-store.yaml

@@ -0,0 +1,278 @@
+version: 2
+image_name: starter
+apis:
+- agents
+- batches
+- datasetio
+- eval
+- files
+- inference
+- post_training
+- safety
+- scoring
+- tool_runtime
+- vector_io
+providers:
+  inference:
+  - provider_id: ${env.CEREBRAS_API_KEY:+cerebras}
+    provider_type: remote::cerebras
+    config:
+      base_url: https://api.cerebras.ai
+      api_key: ${env.CEREBRAS_API_KEY:=}
+  - provider_id: ${env.OLLAMA_URL:+ollama}
+    provider_type: remote::ollama
+    config:
+      url: ${env.OLLAMA_URL:=http://localhost:11434}
+  - provider_id: ${env.VLLM_URL:+vllm}
+    provider_type: remote::vllm
+    config:
+      url: ${env.VLLM_URL:=}
+      max_tokens: ${env.VLLM_MAX_TOKENS:=4096}
+      api_token: ${env.VLLM_API_TOKEN:=fake}
+      tls_verify: ${env.VLLM_TLS_VERIFY:=true}
+  - provider_id: ${env.TGI_URL:+tgi}
+    provider_type: remote::tgi
+    config:
+      url: ${env.TGI_URL:=}
+  - provider_id: fireworks
+    provider_type: remote::fireworks
+    config:
+      url: https://api.fireworks.ai/inference/v1
+      api_key: ${env.FIREWORKS_API_KEY:=}
+  - provider_id: together
+    provider_type: remote::together
+    config:
+      url: https://api.together.xyz/v1
+      api_key: ${env.TOGETHER_API_KEY:=}
+  - provider_id: bedrock
+    provider_type: remote::bedrock
+  - provider_id: ${env.NVIDIA_API_KEY:+nvidia}
+    provider_type: remote::nvidia
+    config:
+      url: ${env.NVIDIA_BASE_URL:=https://integrate.api.nvidia.com}
+      api_key: ${env.NVIDIA_API_KEY:=}
+      append_api_version: ${env.NVIDIA_APPEND_API_VERSION:=True}
+  - provider_id: openai
+    provider_type: remote::openai
+    config:
+      api_key: ${env.OPENAI_API_KEY:=}
+      base_url: ${env.OPENAI_BASE_URL:=https://api.openai.com/v1}
+  - provider_id: anthropic
+    provider_type: remote::anthropic
+    config:
+      api_key: ${env.ANTHROPIC_API_KEY:=}
+  - provider_id: gemini
+    provider_type: remote::gemini
+    config:
+      api_key: ${env.GEMINI_API_KEY:=}
+  - provider_id: ${env.VERTEX_AI_PROJECT:+vertexai}
+    provider_type: remote::vertexai
+    config:
+      project: ${env.VERTEX_AI_PROJECT:=}
+      location: ${env.VERTEX_AI_LOCATION:=us-central1}
+  - provider_id: groq
+    provider_type: remote::groq
+    config:
+      url: https://api.groq.com
+      api_key: ${env.GROQ_API_KEY:=}
+  - provider_id: sambanova
+    provider_type: remote::sambanova
+    config:
+      url: https://api.sambanova.ai/v1
+      api_key: ${env.SAMBANOVA_API_KEY:=}
+  - provider_id: ${env.AZURE_API_KEY:+azure}
+    provider_type: remote::azure
+    config:
+      api_key: ${env.AZURE_API_KEY:=}
+      api_base: ${env.AZURE_API_BASE:=}
+      api_version: ${env.AZURE_API_VERSION:=}
+      api_type: ${env.AZURE_API_TYPE:=}
+  - provider_id: sentence-transformers
+    provider_type: inline::sentence-transformers
+  vector_io:
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      persistence:
+        namespace: vector_io::faiss
+        backend: kv_default
+  - provider_id: sqlite-vec
+    provider_type: inline::sqlite-vec
+    config:
+      db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/starter}/sqlite_vec.db
+      persistence:
+        namespace: vector_io::sqlite_vec
+        backend: kv_default
+  - provider_id: ${env.MILVUS_URL:+milvus}
+    provider_type: inline::milvus
+    config:
+      db_path: ${env.MILVUS_DB_PATH:=~/.llama/distributions/starter}/milvus.db
+      persistence:
+        namespace: vector_io::milvus
+        backend: kv_default
+  - provider_id: ${env.CHROMADB_URL:+chromadb}
+    provider_type: remote::chromadb
+    config:
+      url: ${env.CHROMADB_URL:=}
+      persistence:
+        namespace: vector_io::chroma_remote
+        backend: kv_default
+  - provider_id: ${env.PGVECTOR_DB:+pgvector}
+    provider_type: remote::pgvector
+    config:
+      host: ${env.PGVECTOR_HOST:=localhost}
+      port: ${env.PGVECTOR_PORT:=5432}
+      db: ${env.PGVECTOR_DB:=}
+      user: ${env.PGVECTOR_USER:=}
+      password: ${env.PGVECTOR_PASSWORD:=}
+      persistence:
+        namespace: vector_io::pgvector
+        backend: kv_default
+  - provider_id: ${env.QDRANT_URL:+qdrant}
+    provider_type: remote::qdrant
+    config:
+      api_key: ${env.QDRANT_API_KEY:=}
+      persistence:
+        namespace: vector_io::qdrant_remote
+        backend: kv_default
+  - provider_id: ${env.WEAVIATE_CLUSTER_URL:+weaviate}
+    provider_type: remote::weaviate
+    config:
+      weaviate_api_key: null
+      weaviate_cluster_url: ${env.WEAVIATE_CLUSTER_URL:=localhost:8080}
+      persistence:
+        namespace: vector_io::weaviate
+        backend: kv_default
+  files:
+  - provider_id: meta-reference-files
+    provider_type: inline::localfs
+    config:
+      storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/distributions/starter/files}
+      metadata_store:
+        table_name: files_metadata
+        backend: sql_default
+  safety:
+  - provider_id: llama-guard
+    provider_type: inline::llama-guard
+    config:
+      excluded_categories: []
+  - provider_id: code-scanner
+    provider_type: inline::code-scanner
+  agents:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      persistence_store:
+        type: sql_postgres
+        host: ${env.POSTGRES_HOST:=localhost}
+        port: ${env.POSTGRES_PORT:=5432}
+        db: ${env.POSTGRES_DB:=llamastack}
+        user: ${env.POSTGRES_USER:=llamastack}
+        password: ${env.POSTGRES_PASSWORD:=llamastack}
+      responses_store:
+        type: sql_postgres
+        host: ${env.POSTGRES_HOST:=localhost}
+        port: ${env.POSTGRES_PORT:=5432}
+        db: ${env.POSTGRES_DB:=llamastack}
+        user: ${env.POSTGRES_USER:=llamastack}
+        password: ${env.POSTGRES_PASSWORD:=llamastack}
+  post_training:
+  - provider_id: torchtune-cpu
+    provider_type: inline::torchtune-cpu
+    config:
+      checkpoint_format: meta
+  eval:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      kvstore:
+        namespace: eval
+        backend: kv_default
+  datasetio:
+  - provider_id: huggingface
+    provider_type: remote::huggingface
+    config:
+      kvstore:
+        namespace: datasetio::huggingface
+        backend: kv_default
+  - provider_id: localfs
+    provider_type: inline::localfs
+    config:
+      kvstore:
+        namespace: datasetio::localfs
+        backend: kv_default
+  scoring:
+  - provider_id: basic
+    provider_type: inline::basic
+  - provider_id: llm-as-judge
+    provider_type: inline::llm-as-judge
+  - provider_id: braintrust
+    provider_type: inline::braintrust
+    config:
+      openai_api_key: ${env.OPENAI_API_KEY:=}
+  tool_runtime:
+  - provider_id: brave-search
+    provider_type: remote::brave-search
+    config:
+      api_key: ${env.BRAVE_SEARCH_API_KEY:=}
+      max_results: 3
+  - provider_id: tavily-search
+    provider_type: remote::tavily-search
+    config:
+      api_key: ${env.TAVILY_SEARCH_API_KEY:=}
+      max_results: 3
+  - provider_id: rag-runtime
+    provider_type: inline::rag-runtime
+  - provider_id: model-context-protocol
+    provider_type: remote::model-context-protocol
+  batches:
+  - provider_id: reference
+    provider_type: inline::reference
+    config:
+      kvstore:
+        namespace: batches
+        backend: kv_postgres
+storage:
+  backends:
+    kv_postgres:
+      type: kv_postgres
+      host: ${env.POSTGRES_HOST:=localhost}
+      port: ${env.POSTGRES_PORT:=5432}
+      db: ${env.POSTGRES_DB:=llamastack}
+      user: ${env.POSTGRES_USER:=llamastack}
+      password: ${env.POSTGRES_PASSWORD:=llamastack}
+      table_name: ${env.POSTGRES_TABLE_NAME:=llamastack_kvstore}
+    sql_postgres:
+      type: sql_postgres
+      host: ${env.POSTGRES_HOST:=localhost}
+      port: ${env.POSTGRES_PORT:=5432}
+      db: ${env.POSTGRES_DB:=llamastack}
+      user: ${env.POSTGRES_USER:=llamastack}
+      password: ${env.POSTGRES_PASSWORD:=llamastack}
+  stores:
+    metadata:
+      namespace: registry
+      backend: kv_postgres
+    inference:
+      table_name: inference_store
+      backend: sql_postgres
+      max_write_queue_size: 10000
+      num_writers: 4
+    conversations:
+      table_name: openai_conversations
+      backend: sql_postgres
+    prompts:
+      namespace: prompts
+      backend: kv_postgres
+registered_resources:
+  models: []
+  shields: []
+  vector_dbs: []
+  datasets: []
+  scoring_fns: []
+  benchmarks: []
+  tool_groups: []
+server:
+  port: 8321
+telemetry:
+  enabled: true

src/llama_stack/distributions/starter/starter.py

@@ -17,6 +17,11 @@ from llama_stack.core.datatypes import (
     ToolGroupInput,
     VectorStoresConfig,
 )
+from llama_stack.core.storage.datatypes import (
+    InferenceStoreReference,
+    KVStoreReference,
+    SqlStoreReference,
+)
 from llama_stack.core.utils.dynamic import instantiate_class_type
 from llama_stack.distributions.template import DistributionTemplate, RunConfigSettings
 from llama_stack.providers.datatypes import RemoteProviderSpec
@@ -36,6 +41,7 @@ from llama_stack.providers.remote.vector_io.pgvector.config import (
 )
 from llama_stack.providers.remote.vector_io.qdrant.config import QdrantVectorIOConfig
 from llama_stack.providers.remote.vector_io.weaviate.config import WeaviateVectorIOConfig
+from llama_stack.providers.utils.kvstore.config import PostgresKVStoreConfig
 from llama_stack.providers.utils.sqlstore.sqlstore import PostgresSqlStoreConfig
 
 
@@ -181,6 +187,62 @@ def get_distribution_template(name: str = "starter") -> DistributionTemplate:
             provider_shield_id="${env.CODE_SCANNER_MODEL:=}",
         ),
     ]
+    postgres_config = PostgresSqlStoreConfig.sample_run_config()
+    default_overrides = {
+        "inference": remote_inference_providers + [embedding_provider],
+        "vector_io": [
+            Provider(
+                provider_id="faiss",
+                provider_type="inline::faiss",
+                config=FaissVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
+            ),
+            Provider(
+                provider_id="sqlite-vec",
+                provider_type="inline::sqlite-vec",
+                config=SQLiteVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
+            ),
+            Provider(
+                provider_id="${env.MILVUS_URL:+milvus}",
+                provider_type="inline::milvus",
+                config=MilvusVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
+            ),
+            Provider(
+                provider_id="${env.CHROMADB_URL:+chromadb}",
+                provider_type="remote::chromadb",
+                config=ChromaVectorIOConfig.sample_run_config(
+                    f"~/.llama/distributions/{name}/",
+                    url="${env.CHROMADB_URL:=}",
+                ),
+            ),
+            Provider(
+                provider_id="${env.PGVECTOR_DB:+pgvector}",
+                provider_type="remote::pgvector",
+                config=PGVectorVectorIOConfig.sample_run_config(
+                    f"~/.llama/distributions/{name}",
+                    db="${env.PGVECTOR_DB:=}",
+                    user="${env.PGVECTOR_USER:=}",
+                    password="${env.PGVECTOR_PASSWORD:=}",
+                ),
+            ),
+            Provider(
+                provider_id="${env.QDRANT_URL:+qdrant}",
+                provider_type="remote::qdrant",
+                config=QdrantVectorIOConfig.sample_run_config(
+                    f"~/.llama/distributions/{name}",
+                    url="${env.QDRANT_URL:=}",
+                ),
+            ),
+            Provider(
+                provider_id="${env.WEAVIATE_CLUSTER_URL:+weaviate}",
+                provider_type="remote::weaviate",
+                config=WeaviateVectorIOConfig.sample_run_config(
+                    f"~/.llama/distributions/{name}",
+                    cluster_url="${env.WEAVIATE_CLUSTER_URL:=}",
+                ),
+            ),
+        ],
+        "files": [files_provider],
+    }
 
     return DistributionTemplate(
         name=name,
@@ -189,64 +251,10 @@ def get_distribution_template(name: str = "starter") -> DistributionTemplate:
         container_image=None,
         template_path=None,
         providers=providers,
-        additional_pip_packages=PostgresSqlStoreConfig.pip_packages(),
+        additional_pip_packages=list(set(PostgresSqlStoreConfig.pip_packages() + PostgresKVStoreConfig.pip_packages())),
         run_configs={
             "run.yaml": RunConfigSettings(
-                provider_overrides={
+                provider_overrides=default_overrides,
-                    "inference": remote_inference_providers + [embedding_provider],
-                    "vector_io": [
-                        Provider(
-                            provider_id="faiss",
-                            provider_type="inline::faiss",
-                            config=FaissVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
-                        ),
-                        Provider(
-                            provider_id="sqlite-vec",
-                            provider_type="inline::sqlite-vec",
-                            config=SQLiteVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
-                        ),
-                        Provider(
-                            provider_id="${env.MILVUS_URL:+milvus}",
-                            provider_type="inline::milvus",
-                            config=MilvusVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
-                        ),
-                        Provider(
-                            provider_id="${env.CHROMADB_URL:+chromadb}",
-                            provider_type="remote::chromadb",
-                            config=ChromaVectorIOConfig.sample_run_config(
-                                f"~/.llama/distributions/{name}/",
-                                url="${env.CHROMADB_URL:=}",
-                            ),
-                        ),
-                        Provider(
-                            provider_id="${env.PGVECTOR_DB:+pgvector}",
-                            provider_type="remote::pgvector",
-                            config=PGVectorVectorIOConfig.sample_run_config(
-                                f"~/.llama/distributions/{name}",
-                                db="${env.PGVECTOR_DB:=}",
-                                user="${env.PGVECTOR_USER:=}",
-                                password="${env.PGVECTOR_PASSWORD:=}",
-                            ),
-                        ),
-                        Provider(
-                            provider_id="${env.QDRANT_URL:+qdrant}",
-                            provider_type="remote::qdrant",
-                            config=QdrantVectorIOConfig.sample_run_config(
-                                f"~/.llama/distributions/{name}",
-                                url="${env.QDRANT_URL:=}",
-                            ),
-                        ),
-                        Provider(
-                            provider_id="${env.WEAVIATE_CLUSTER_URL:+weaviate}",
-                            provider_type="remote::weaviate",
-                            config=WeaviateVectorIOConfig.sample_run_config(
-                                f"~/.llama/distributions/{name}",
-                                cluster_url="${env.WEAVIATE_CLUSTER_URL:=}",
-                            ),
-                        ),
-                    ],
-                    "files": [files_provider],
-                },
                 default_models=[],
                 default_tool_groups=default_tool_groups,
                 default_shields=default_shields,
@@ -261,6 +269,55 @@ def get_distribution_template(name: str = "starter") -> DistributionTemplate:
                     default_shield_id="llama-guard",
                 ),
             ),
+            "run-with-postgres-store.yaml": RunConfigSettings(
+                provider_overrides={
+                    **default_overrides,
+                    "agents": [
+                        Provider(
+                            provider_id="meta-reference",
+                            provider_type="inline::meta-reference",
+                            config=dict(
+                                persistence_store=postgres_config,
+                                responses_store=postgres_config,
+                            ),
+                        )
+                    ],
+                    "batches": [
+                        Provider(
+                            provider_id="reference",
+                            provider_type="inline::reference",
+                            config=dict(
+                                kvstore=KVStoreReference(
+                                    backend="kv_postgres",
+                                    namespace="batches",
+                                ).model_dump(exclude_none=True),
+                            ),
+                        )
+                    ],
+                },
+                storage_backends={
+                    "kv_postgres": PostgresKVStoreConfig.sample_run_config(),
+                    "sql_postgres": postgres_config,
+                },
+                storage_stores={
+                    "metadata": KVStoreReference(
+                        backend="kv_postgres",
+                        namespace="registry",
+                    ).model_dump(exclude_none=True),
+                    "inference": InferenceStoreReference(
+                        backend="sql_postgres",
+                        table_name="inference_store",
+                    ).model_dump(exclude_none=True),
+                    "conversations": SqlStoreReference(
+                        backend="sql_postgres",
+                        table_name="openai_conversations",
+                    ).model_dump(exclude_none=True),
+                    "prompts": KVStoreReference(
+                        backend="kv_postgres",
+                        namespace="prompts",
+                    ).model_dump(exclude_none=True),
+                },
+            ),
         },
         run_config_env_vars={
             "LLAMA_STACK_PORT": (

src/llama_stack/providers/remote/inference/passthrough/__init__.py

@@ -10,8 +10,8 @@ from .config import PassthroughImplConfig
 
 
 class PassthroughProviderDataValidator(BaseModel):
-    url: str
+    passthrough_url: str
-    api_key: str
+    passthrough_api_key: str
 
 
 async def get_adapter_impl(config: PassthroughImplConfig, _deps):

src/llama_stack/providers/remote/inference/passthrough/config.py

@@ -6,7 +6,7 @@
 
 from typing import Any
 
-from pydantic import Field, SecretStr
+from pydantic import Field
 
 from llama_stack.providers.utils.inference.model_registry import RemoteInferenceProviderConfig
 from llama_stack.schema_utils import json_schema_type
@@ -19,11 +19,6 @@ class PassthroughImplConfig(RemoteInferenceProviderConfig):
         description="The URL for the passthrough endpoint",
     )
 
-    api_key: SecretStr | None = Field(
-        default=None,
-        description="API Key for the passthrouth endpoint",
-    )
-
     @classmethod
     def sample_run_config(
         cls, url: str = "${env.PASSTHROUGH_URL}", api_key: str = "${env.PASSTHROUGH_API_KEY}", **kwargs

src/llama_stack/providers/remote/inference/passthrough/passthrough.py

@@ -5,9 +5,8 @@
 # the root directory of this source tree.
 
 from collections.abc import AsyncIterator
-from typing import Any
 
-from llama_stack_client import AsyncLlamaStackClient
+from openai import AsyncOpenAI
 
 from llama_stack.apis.inference import (
     Inference,
@@ -20,103 +19,117 @@ from llama_stack.apis.inference import (
     OpenAIEmbeddingsResponse,
 )
 from llama_stack.apis.models import Model
-from llama_stack.core.library_client import convert_pydantic_to_json_value
+from llama_stack.core.request_headers import NeedsRequestProviderData
-from llama_stack.providers.utils.inference.model_registry import ModelRegistryHelper
 
 from .config import PassthroughImplConfig
 
 
-class PassthroughInferenceAdapter(Inference):
+class PassthroughInferenceAdapter(NeedsRequestProviderData, Inference):
     def __init__(self, config: PassthroughImplConfig) -> None:
-        ModelRegistryHelper.__init__(self)
         self.config = config
 
+    async def initialize(self) -> None:
+        pass
+
+    async def shutdown(self) -> None:
+        pass
+
     async def unregister_model(self, model_id: str) -> None:
         pass
 
     async def register_model(self, model: Model) -> Model:
         return model
 
-    def _get_client(self) -> AsyncLlamaStackClient:
+    async def list_models(self) -> list[Model]:
-        passthrough_url = None
+        """List models by calling the downstream /v1/models endpoint."""
-        passthrough_api_key = None
+        client = self._get_openai_client()
-        provider_data = None
 
-        if self.config.url is not None:
+        response = await client.models.list()
-            passthrough_url = self.config.url
-        else:
-            provider_data = self.get_request_provider_data()
-            if provider_data is None or not provider_data.passthrough_url:
-                raise ValueError(
-                    'Pass url of the passthrough endpoint in the header X-LlamaStack-Provider-Data as { "passthrough_url": <your passthrough url>}'
-                )
-            passthrough_url = provider_data.passthrough_url
 
-        if self.config.api_key is not None:
+        # Convert from OpenAI format to Llama Stack Model format
-            passthrough_api_key = self.config.api_key.get_secret_value()
+        models = []
-        else:
+        for model_data in response.data:
-            provider_data = self.get_request_provider_data()
+            downstream_model_id = model_data.id
-            if provider_data is None or not provider_data.passthrough_api_key:
+            custom_metadata = getattr(model_data, "custom_metadata", {}) or {}
-                raise ValueError(
-                    'Pass API Key for the passthrough endpoint in the header X-LlamaStack-Provider-Data as { "passthrough_api_key": <your api key>}'
-                )
-            passthrough_api_key = provider_data.passthrough_api_key
 
-        return AsyncLlamaStackClient(
+            # Prefix identifier with provider ID for local registry
-            base_url=passthrough_url,
+            local_identifier = f"{self.__provider_id__}/{downstream_model_id}"
-            api_key=passthrough_api_key,
+
-            provider_data=provider_data,
+            model = Model(
+                identifier=local_identifier,
+                provider_id=self.__provider_id__,
+                provider_resource_id=downstream_model_id,
+                model_type=custom_metadata.get("model_type", "llm"),
+                metadata=custom_metadata,
+            )
+            models.append(model)
+
+        return models
+
+    async def should_refresh_models(self) -> bool:
+        """Passthrough should refresh models since they come from downstream dynamically."""
+        return self.config.refresh_models
+
+    def _get_openai_client(self) -> AsyncOpenAI:
+        """Get an AsyncOpenAI client configured for the downstream server."""
+        base_url = self._get_passthrough_url()
+        api_key = self._get_passthrough_api_key()
+
+        return AsyncOpenAI(
+            base_url=f"{base_url.rstrip('/')}/v1",
+            api_key=api_key,
         )
 
-    async def openai_embeddings(
+    def _get_passthrough_url(self) -> str:
-        self,
+        """Get the passthrough URL from config or provider data."""
-        params: OpenAIEmbeddingsRequestWithExtraBody,
+        if self.config.url is not None:
-    ) -> OpenAIEmbeddingsResponse:
+            return self.config.url
-        raise NotImplementedError()
+
+        provider_data = self.get_request_provider_data()
+        if provider_data is None:
+            raise ValueError(
+                'Pass url of the passthrough endpoint in the header X-LlamaStack-Provider-Data as { "passthrough_url": <your passthrough url>}'
+            )
+        return provider_data.passthrough_url
+
+    def _get_passthrough_api_key(self) -> str:
+        """Get the passthrough API key from config or provider data."""
+        if self.config.auth_credential is not None:
+            return self.config.auth_credential.get_secret_value()
+
+        provider_data = self.get_request_provider_data()
+        if provider_data is None:
+            raise ValueError(
+                'Pass API Key for the passthrough endpoint in the header X-LlamaStack-Provider-Data as { "passthrough_api_key": <your api key>}'
+            )
+        return provider_data.passthrough_api_key
 
     async def openai_completion(
         self,
         params: OpenAICompletionRequestWithExtraBody,
     ) -> OpenAICompletion:
-        client = self._get_client()
+        """Forward completion request to downstream using OpenAI client."""
-        model_obj = await self.model_store.get_model(params.model)
+        client = self._get_openai_client()
-
-        params = params.model_copy()
-        params.model = model_obj.provider_resource_id
-
         request_params = params.model_dump(exclude_none=True)
-
+        response = await client.completions.create(**request_params)
-        return await client.inference.openai_completion(**request_params)
+        return response  # type: ignore
 
     async def openai_chat_completion(
         self,
         params: OpenAIChatCompletionRequestWithExtraBody,
     ) -> OpenAIChatCompletion | AsyncIterator[OpenAIChatCompletionChunk]:
-        client = self._get_client()
+        """Forward chat completion request to downstream using OpenAI client."""
-        model_obj = await self.model_store.get_model(params.model)
+        client = self._get_openai_client()
-
-        params = params.model_copy()
-        params.model = model_obj.provider_resource_id
-
         request_params = params.model_dump(exclude_none=True)
+        response = await client.chat.completions.create(**request_params)
+        return response  # type: ignore
 
-        return await client.inference.openai_chat_completion(**request_params)
+    async def openai_embeddings(
-
+        self,
-    def cast_value_to_json_dict(self, request_params: dict[str, Any]) -> dict[str, Any]:
+        params: OpenAIEmbeddingsRequestWithExtraBody,
-        json_params = {}
+    ) -> OpenAIEmbeddingsResponse:
-        for key, value in request_params.items():
+        """Forward embeddings request to downstream using OpenAI client."""
-            json_input = convert_pydantic_to_json_value(value)
+        client = self._get_openai_client()
-            if isinstance(json_input, dict):
+        request_params = params.model_dump(exclude_none=True)
-                json_input = {k: v for k, v in json_input.items() if v is not None}
+        response = await client.embeddings.create(**request_params)
-            elif isinstance(json_input, list):
+        return response  # type: ignore
-                json_input = [x for x in json_input if x is not None]
-                new_input = []
-                for x in json_input:
-                    if isinstance(x, dict):
-                        x = {k: v for k, v in x.items() if v is not None}
-                    new_input.append(x)
-                json_input = new_input
-
-            json_params[key] = json_input
-
-        return json_params

src/llama_stack/providers/utils/memory/openai_vector_store_mixin.py

@@ -26,6 +26,7 @@ from llama_stack.apis.vector_io import (
     VectorStoreChunkingStrategy,
     VectorStoreChunkingStrategyAuto,
     VectorStoreChunkingStrategyStatic,
+    VectorStoreChunkingStrategyStaticConfig,
     VectorStoreContent,
     VectorStoreDeleteResponse,
     VectorStoreFileBatchObject,
@@ -414,6 +415,10 @@ class OpenAIVectorStoreMixin(ABC):
             in_progress=0,
             total=0,
         )
+        if not params.chunking_strategy or params.chunking_strategy.type == "auto":
+            chunking_strategy = VectorStoreChunkingStrategyStatic(static=VectorStoreChunkingStrategyStaticConfig())
+        else:
+            chunking_strategy = params.chunking_strategy
         store_info: dict[str, Any] = {
             "id": vector_store_id,
             "object": "vector_store",
@@ -426,7 +431,7 @@ class OpenAIVectorStoreMixin(ABC):
             "expires_at": None,
             "last_active_at": created_at,
             "file_ids": [],
-            "chunking_strategy": params.chunking_strategy,
+            "chunking_strategy": chunking_strategy.model_dump(),
         }
 
         # Add provider information to metadata if provided
@@ -637,7 +642,7 @@ class OpenAIVectorStoreMixin(ABC):
                     break
 
             return VectorStoreSearchResponsePage(
-                search_query=search_query,
+                search_query=query if isinstance(query, list) else [query],
                 data=data,
                 has_more=False,  # For simplicity, we don't implement pagination here
                 next_page=None,
@@ -647,7 +652,7 @@ class OpenAIVectorStoreMixin(ABC):
             logger.error(f"Error searching vector store {vector_store_id}: {e}")
             # Return empty results on error
             return VectorStoreSearchResponsePage(
-                search_query=search_query,
+                search_query=query if isinstance(query, list) else [query],
                 data=[],
                 has_more=False,
                 next_page=None,
@@ -886,8 +891,8 @@ class OpenAIVectorStoreMixin(ABC):
 
         # Determine pagination info
         has_more = len(file_objects) > limit
-        first_id = file_objects[0].id if file_objects else None
+        first_id = limited_files[0].id if file_objects else None
-        last_id = file_objects[-1].id if file_objects else None
+        last_id = limited_files[-1].id if file_objects else None
 
         return VectorStoreListFilesResponse(
             data=limited_files,

tests/integration/ci_matrix.json

@@ -2,7 +2,8 @@
   "default": [
     {"suite": "base", "setup": "ollama"},
     {"suite": "vision", "setup": "ollama-vision"},
-    {"suite": "responses", "setup": "gpt"}
+    {"suite": "responses", "setup": "gpt"},
+    {"suite": "base-vllm-subset", "setup": "vllm"}
   ],
   "schedules": {
     "1 0 * * 0": [

tests/integration/common/recordings/models-64a2277c90f0f42576f60c1030e3a020403d34a95f56931b792d5939f4cebc57-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": null,
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374291,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-f70298e4ea3e4b4eb7f2cc2deb7a2b01",
+              "object": "model_permission",
+              "created": 1762374291,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/452805c3c85951c86e4e5dfeef078a2e184866dafee83186cd84932daae1af42.json

@@ -0,0 +1,84 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestEdgeCases::test_tool_without_schema[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Call the no args tool"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "no_args_tool",
+            "description": "Tool with no arguments",
+            "parameters": {
+              "type": "object",
+              "properties": {}
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-452805c3c859",
+        "choices": [
+          {
+            "finish_reason": "tool_calls",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user wants me to call the no args tool. Let me check the available functions. There's only one tool provided, which is the no_args_tool with no arguments. Since the user didn't specify any parameters, I should just return the tool call as instructed. I need to make sure the JSON is correctly formatted and within the XML tags. Alright, that's all I need.\n</think>\n\n",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [
+                {
+                  "id": "chatcmpl-tool-7a67269afe214c85924c5171612bbdbd",
+                  "function": {
+                    "arguments": "{}",
+                    "name": "no_args_tool"
+                  },
+                  "type": "function"
+                }
+              ],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 101,
+          "prompt_tokens": 136,
+          "total_tokens": 237,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/496035259763c1bddb1a3148c2586663d08a5bc31f697d1fc5d9bed1c71f5950.json

@@ -0,0 +1,92 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestOpenAICompatibility::test_openai_chat_completion_with_tools[openai_client-txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "What's the weather in Tokyo?"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "get_weather",
+            "description": "Get weather information",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "location": {
+                  "type": "string",
+                  "description": "City name"
+                }
+              },
+              "required": [
+                "location"
+              ]
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-496035259763",
+        "choices": [
+          {
+            "finish_reason": "tool_calls",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user is asking about the weather in Tokyo. I need to use the get_weather function for that. The function requires the location parameter, which in this case is Tokyo. I should make sure to specify \"Tokyo\" as the location. Let me check if there are any other parameters needed, but no, the function only needs the location. So the tool call should be straightforward. I'll format the JSON correctly inside the tool_call tags.\n</think>\n\n",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [
+                {
+                  "id": "chatcmpl-tool-959b557fa67e4134a2391f5d35e5d5ae",
+                  "function": {
+                    "arguments": "{\"location\": \"Tokyo\"}",
+                    "name": "get_weather"
+                  },
+                  "type": "function"
+                }
+              ],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 117,
+          "prompt_tokens": 158,
+          "total_tokens": 275,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/524ead18daaddb6228284820adaa3fb312d2a525cc35e20c181190ddf40793e6.json

@@ -0,0 +1,92 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestOpenAICompatibility::test_openai_format_preserves_complex_schemas[openai_client-txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Process this data"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "process_data",
+            "description": "Process structured data",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "data": {
+                  "$ref": "#/$defs/DataObject"
+                }
+              },
+              "$defs": {
+                "DataObject": {
+                  "type": "object",
+                  "properties": {
+                    "values": {
+                      "type": "array",
+                      "items": {
+                        "type": "number"
+                      }
+                    }
+                  }
+                }
+              }
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-524ead18daad",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user wants me to process the data. Let me check the available tools. There's a function called process_data that takes an object with a 'data' parameter. The data is an array of numbers. But the user hasn't provided any specific data yet. They just said \"Process this data.\" Hmm, maybe they expect me to prompt them for the data first. Wait, maybe there's a misunderstanding. Did they include the data in the conversation history? Let me look back. The user's message is \"Process this data.\" No data provided. Oh, maybe they made a mistake and forgot to include it. I need to ask them to provide the data so I can proceed. Let me confirm if there's any data mentioned. No, the current input is just the instruction. So I should ask the user to supply the data array of numbers to process.\n</think>\n\nPlease provide the structured data you'd like me to process. For example, an array of numbers like `[1, 2, 3]`.",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 212,
+          "prompt_tokens": 180,
+          "total_tokens": 392,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/744052775cf90e30dac587e6b809d41a8cc37adc29c500eecee2727f428cbf5a.json

@@ -0,0 +1,98 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_with_tools[txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:tool_calling]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "system",
+          "content": "Pretend you are a weather assistant."
+        },
+        {
+          "role": "user",
+          "content": "What's the weather like in San Francisco, CA?"
+        }
+      ],
+      "max_tokens": 4096,
+      "stream": false,
+      "tool_choice": "auto",
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "get_weather",
+            "description": "Get the current weather",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "location": {
+                  "type": "string",
+                  "description": "The city and state (both required), e.g. San Francisco, CA."
+                }
+              },
+              "required": [
+                "location"
+              ]
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-744052775cf9",
+        "choices": [
+          {
+            "finish_reason": "tool_calls",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user is asking about the weather in San Francisco, CA. I need to use the get_weather function. The function requires the location parameter, which is provided as San Francisco, CA. I should make sure to format the arguments correctly as a JSON object. Let me check the required parameters again. The location is required, so I can't omit it. I'll structure the tool call with the name \"get_weather\" and the arguments including \"location\": \"San Francisco, CA\". That should get the current weather information for the user.\n</think>\n\n",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [
+                {
+                  "id": "chatcmpl-tool-b59dc311dd914d3dbd6d455b122bc39c",
+                  "function": {
+                    "arguments": "{\"location\": \"San Francisco, CA\"}",
+                    "name": "get_weather"
+                  },
+                  "type": "function"
+                }
+              ],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 138,
+          "prompt_tokens": 185,
+          "total_tokens": 323,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/77cf218283607bfac37623e1bb4e6f33cae670df7d6995d432bca34c5dfb0e43.json

@@ -0,0 +1,67 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_with_tool_choice_none[txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:tool_calling]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "system",
+          "content": "Pretend you are a weather assistant."
+        },
+        {
+          "role": "user",
+          "content": "What's the weather like in San Francisco, CA?"
+        }
+      ],
+      "max_tokens": 4096,
+      "stream": false
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-77cf21828360",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user is asking about the weather in San Francisco, CA. I need to check the current weather conditions. But wait, I can't access real-time data. I should mention that I can't provide the current weather forecast and ask them to check a reliable source like the National Weather Service or a weather app. Also, maybe suggest they can provide more details if they need help with something else related to the weather.\n</think>\n\nI'm sorry, but I can't provide real-time weather information. However, you can check the current weather for San Francisco, CA using the National Weather Service (NWS) website, weather apps like Weather.com, or local meteorological services. Let me know if there's anything else I can assist with!",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 154,
+          "prompt_tokens": 33,
+          "total_tokens": 187,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/853f6a700b98d71d390b7d366e27133a22772fbdf11863158349c1b0625bbc72.json

@@ -0,0 +1,128 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestEdgeCases::test_multiple_tools_with_different_schemas[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Use one of the available tools"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "simple",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "x": {
+                  "type": "string"
+                }
+              }
+            }
+          }
+        },
+        {
+          "type": "function",
+          "function": {
+            "name": "complex",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "data": {
+                  "$ref": "#/$defs/Complex"
+                }
+              },
+              "$defs": {
+                "Complex": {
+                  "type": "object",
+                  "properties": {
+                    "nested": {
+                      "type": "array",
+                      "items": {
+                        "type": "number"
+                      }
+                    }
+                  }
+                }
+              }
+            }
+          }
+        },
+        {
+          "type": "function",
+          "function": {
+            "name": "with_output",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "input": {
+                  "type": "string"
+                }
+              }
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-853f6a700b98",
+        "choices": [
+          {
+            "finish_reason": "tool_calls",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, let's see. The user wants me to use one of the available tools. The tools provided are simple, complex, and with_output. The simple function takes an argument 'x' of type string. The complex function requires a 'data' parameter that's an object with a nested array of numbers. The with_output function takes an input string.\n\nThe user's query is about using a tool, but there's no specific function name mentioned. Wait, maybe the user expects me to choose one based on the context. Since the tools are available, but the query is a general instruction, perhaps I should ask for clarification. However, the instructions say to use one of the tools if possible. Since the user hasn't specified a particular function, maybe I should check if there's any implicit function needed. But looking at the tools, none are directly related to the query. The user might need to specify which tool to use. Alternatively, maybe the answer is to call the simple function with an example input. But without more context, it's hard to tell. Wait, maybe the user expects me to choose the simplest one. Let's go with the simple function first. So the tool call would be to the simple function with x set to some value. But the user hasn't provided a specific value. Maybe I should state that the tool requires a value. But according to the instructions, if possible, use one of the tools. Since the user hasn't given a value, perhaps the answer is to call the simple function with an example. But the parameters for the simple function require 'x' which is a string. Maybe the user expects me to proceed without needing more info. So I'll proceed by calling the simple function with x as \"example\".\n</think>\n\n",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [
+                {
+                  "id": "chatcmpl-tool-12e2ba0189cf484bb936cbb254a5c32a",
+                  "function": {
+                    "arguments": "{\"x\": \"example\"}",
+                    "name": "simple"
+                  },
+                  "type": "function"
+                }
+              ],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 378,
+          "prompt_tokens": 265,
+          "total_tokens": 643,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/cfb292c0f41dbc4a2c0fb39016760f6c23c248a7bbffea19ac6bcab7bf25292d.json

@@ -0,0 +1,114 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestChatCompletionWithTools::test_tool_with_complex_schema[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Book a flight from SFO to JFK for John Doe"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "book_flight",
+            "description": "Book a flight",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "flight": {
+                  "$ref": "#/$defs/FlightInfo"
+                },
+                "passenger": {
+                  "$ref": "#/$defs/Passenger"
+                }
+              },
+              "required": [
+                "flight",
+                "passenger"
+              ],
+              "$defs": {
+                "FlightInfo": {
+                  "type": "object",
+                  "properties": {
+                    "from": {
+                      "type": "string"
+                    },
+                    "to": {
+                      "type": "string"
+                    },
+                    "date": {
+                      "type": "string",
+                      "format": "date"
+                    }
+                  }
+                },
+                "Passenger": {
+                  "type": "object",
+                  "properties": {
+                    "name": {
+                      "type": "string"
+                    },
+                    "age": {
+                      "type": "integer"
+                    }
+                  }
+                }
+              }
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-cfb292c0f41d",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user wants to book a flight from SFO to JFK for John Doe. Let me check the tools available. The provided function is book_flight, which requires flight information and a passenger. The parameters needed are flight (as a FlightInfo object) and passenger (with name and age). The user mentioned SFO to JFK, so the flight details are from and to. The passenger's name is John Doe, but the age isn't provided. Wait, the function parameters require the passenger's name and age, but the user only mentioned the name. Maybe the age is missing? But the user didn't specify it, so perhaps I should note that the age is required. However, the function's required parameters are flight and passenger, so even if age is missing, the function can't be called without it. So I need to include both flight info and passenger details. The user's message only gives the name and destination, not the flight details or age. Therefore, I need to ask for the flight details and the passenger's age. But the user hasn't provided those. So I can't proceed with the function call. Wait, but maybe the user expects me to assume some default? No, the function requires all parameters. Since the user hasn't provided flight details or age, I can't call the function. So the correct response is to prompt the user for those details.\n</think>\n\nThe booking requires the flight details and passenger's age. Could you provide the flight number and John Doe's age?",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 310,
+          "prompt_tokens": 261,
+          "total_tokens": 571,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/df353403c7fb59ed88c52269261b3dd9b75f681f8bb5431b4f07006d6c08aa7c.json

@@ -0,0 +1,96 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_structured_output[txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:structured_output]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "system",
+          "content": "You are a helpful assistant. Michael Jordan was born in 1963. His first name is \"Michael\", He played basketball for the Chicago Bulls for 15 seasons and was drafted in 1984"
+        },
+        {
+          "role": "user",
+          "content": "Please give me information about Michael Jordan."
+        }
+      ],
+      "max_tokens": 4096,
+      "response_format": {
+        "type": "json_schema",
+        "json_schema": {
+          "name": "AnswerFormat",
+          "schema": {
+            "properties": {
+              "first_name": {
+                "title": "First Name",
+                "type": "string"
+              },
+              "last_name": {
+                "title": "Last Name",
+                "type": "string"
+              },
+              "year_of_birth": {
+                "title": "Year Of Birth",
+                "type": "integer"
+              }
+            },
+            "required": [
+              "first_name",
+              "last_name",
+              "year_of_birth"
+            ],
+            "title": "AnswerFormat",
+            "type": "object"
+          }
+        }
+      },
+      "stream": false
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-df353403c7fb",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "{\"first_name\": \"Michael\", \"last_name\": \"Jordan\", \"year_of_birth\": 1963}",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 28,
+          "prompt_tokens": 66,
+          "total_tokens": 94,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/e89112e7735fccc5ad9ebe6a96454953aed0ba2501cabfaa80b742c2bf371cbc.json

@@ -0,0 +1,92 @@
+{
+  "test_id": "tests/integration/inference/test_tools_with_schemas.py::TestChatCompletionWithTools::test_simple_tool_call[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "messages": [
+        {
+          "role": "user",
+          "content": "What's the weather in San Francisco?"
+        }
+      ],
+      "max_tokens": 4096,
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "name": "get_weather",
+            "description": "Get weather for a location",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "location": {
+                  "type": "string",
+                  "description": "City name"
+                }
+              },
+              "required": [
+                "location"
+              ]
+            }
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-e89112e7735f",
+        "choices": [
+          {
+            "finish_reason": "tool_calls",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "<think>\nOkay, the user is asking for the weather in San Francisco. I need to check if there's a function available for that. Looking at the tools provided, there's a function called get_weather that requires a location parameter. The description says it gets weather for a location, and the parameter is the city name. The user provided \"San Francisco\" as the location, so I should call the get_weather function with \"San Francisco\" as the argument. I don't see any other parameters needed here, so the tool call should be straightforward. Just make sure the city name is correctly formatted in JSON.\n</think>\n\n",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": [
+                {
+                  "id": "chatcmpl-tool-feead29842dc40b2831c41ed397f555f",
+                  "function": {
+                    "arguments": "{\"location\": \"San Francisco\"}",
+                    "name": "get_weather"
+                  },
+                  "type": "function"
+                }
+              ],
+              "reasoning_content": null
+            },
+            "stop_reason": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 146,
+          "prompt_tokens": 161,
+          "total_tokens": 307,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "prompt_logprobs": null,
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/f02f1bfd75adaea87b91dedc59430b99015b5ed0e2bbf24418a31146ffcbca9b.json

@@ -0,0 +1,53 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_guided_choice[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/completions",
+    "headers": {},
+    "body": {
+      "model": "Qwen/Qwen3-0.6B",
+      "prompt": "I am feeling really sad today.",
+      "stream": false,
+      "extra_body": {
+        "guided_choice": [
+          "joy",
+          "sadness"
+        ]
+      }
+    },
+    "endpoint": "/v1/completions",
+    "model": "Qwen/Qwen3-0.6B"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.completion.Completion",
+      "__data__": {
+        "id": "rec-f02f1bfd75ad",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "text": "joy",
+            "stop_reason": null,
+            "prompt_logprobs": null
+          }
+        ],
+        "created": 0,
+        "model": "Qwen/Qwen3-0.6B",
+        "object": "text_completion",
+        "system_fingerprint": null,
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 7,
+          "total_tokens": 9,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        },
+        "kv_transfer_params": null
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-0037f2d2065a360cfcc36c35f138318cfc6508e743ff9423da4b7b1d7bfd4f3f-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_different_inputs_different_outputs[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375180,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-aeeb49e5e51c42fa94562780165bd620",
+              "object": "model_permission",
+              "created": 1762375180,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-009fb75503cf565d6c97f70deb8235432b0020b93d55e3b33ea093664c4bbc82-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_dimensions[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375115,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-feec0a894be04f738e12b596ff163b64",
+              "object": "model_permission",
+              "created": 1762375115,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-01e6ee9852f532d9b0d82dde2e7c831d698e81dea1be69433050d42643f35edc-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_stop_sequence[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:stop_sequence]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374330,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-119e17052e4c4c13bd791af3138d5360",
+              "object": "model_permission",
+              "created": 1762374330,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-10370bf5307b2fc971b8e53bdcc4e9eb4d3d76fe8ecdb31231b59576a612e972-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_base64_batch_processing[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375226,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-c6ae673fda084519b3c67947896cd3b0",
+              "object": "model_permission",
+              "created": 1762375226,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-1312e0d8579e9b0e6dcb222272de34115277db71c6c560872fa13722197f881f-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_encoding_format_base64[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374573,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-3f422354a81e491b87f93d5b192a0e1a",
+              "object": "model_permission",
+              "created": 1762374573,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-134e731d073e9e07eb9782bbe292167f8ad08157a15150ce92135854d04050fc-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming_with_n[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374305,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-794e16e59ddb4216a8bedfdf485b8f24",
+              "object": "model_permission",
+              "created": 1762374305,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-14c0905df1b177d2f85b30b0285b0ffdc88d1a7b290e2155fb7a01f3c1436ca0-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store_tool_calls[openai_client-txt=vllm/Qwen/Qwen3-0.6B-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374317,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-ff7d26d076eb4373a0631a80fe3ae063",
+              "object": "model_permission",
+              "created": 1762374317,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-1bc879637162ba23badeea66c4c25a638869a3e90d16ef3e84dea1a613e7192e-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375033,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-2a16fede981b43be9e1cbe3dbedd1e74",
+              "object": "model_permission",
+              "created": 1762375033,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-2b9bac5da1a03c0b572bc019cc0c50904d49e6193990ca245908f4535bcaab43-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_guided_choice[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374297,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-4bc93704559a4e1d8492aeec7222040c",
+              "object": "model_permission",
+              "created": 1762374297,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-394c30370fe5b724c5fe1292984373b281d47b2ac0d49e8b598f13cf100b3ad8-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_invalid_model_error[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374532,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-e353aa079d5145c19953791ac99daeba",
+              "object": "model_permission",
+              "created": 1762374532,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-3f4208962fdb2be3e7057777fc93a149890bd1dfa8a92597e176f23658e86cd8-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store_tool_calls[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375260,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-10c27d1c9e324b18b65321b422e19af9",
+              "object": "model_permission",
+              "created": 1762375260,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-4a729b00af209ad60846d1904e5973ad081aa5f595de50f5ef1aae304cb67ef3-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming_with_n[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375040,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-f01c211577294936958dd28046c89dba",
+              "object": "model_permission",
+              "created": 1762375040,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-4f62bcb9cdf74f4c2ed804038def162f18ad384182b0f174918607e9ed3c1515-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_non_streaming[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:non_streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375266,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-7166a6fcd331435eb2d0f0a6b23382ed",
+              "object": "model_permission",
+              "created": 1762375266,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-54b420cdb98a0149a618088f55746e26b7bf6e7c5ebf5fa07c13ec9e366521d3-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374301,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-cd16b092c5a04e719ddf786f0c3e935e",
+              "object": "model_permission",
+              "created": 1762374301,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-613f5d11a8cda7126115f96650334fde0a0457a6b4a2605bc15eec9b50a6956c-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_non_streaming[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:sanity]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374295,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-9f71adbb206846bb9d0e12834e41551e",
+              "object": "model_permission",
+              "created": 1762374295,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-62a361f55d61a98ea0863e9acfb5ab5d540c5d19e791415ee476474f7f1ed90f-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_logprobs_streaming[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:log_probs]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374342,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-a8b7b38c40584a03b4b346b6c181fb93",
+              "object": "model_permission",
+              "created": 1762374342,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-668fc72f70ac72d5c112fe79d86d5c790611456b3f0102832f27e6edd420ab54-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_non_streaming[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:non_streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375235,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-dd48560646f141298f5cc2ef3467e54b",
+              "object": "model_permission",
+              "created": 1762375235,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-702eee4572e9b17ff0b0fdd55b10021f7077f0afcba922d6a53db0b537542518-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_user_parameter[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374500,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-0ba0c3a54dcb4e57bc0308fd54425933",
+              "object": "model_permission",
+              "created": 1762374500,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-723d37a5bceab199cff076a0dcc2d4ee7596b7c800f13c64f6a6ecdbf4ed2f3a-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store[openai_client-txt=vllm/Qwen/Qwen3-0.6B-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374311,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-e95a9ed7439245b5995add97fb50f765",
+              "object": "model_permission",
+              "created": 1762374311,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-768c497339830cf86ddd7843f33d0ed06b3bce3ef2ae9f854364b534ba8cafb7-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_encoding_format_float[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375099,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-6b5eba46536f43df902871dd257e1676",
+              "object": "model_permission",
+              "created": 1762375099,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-7b0f2493d699e58cdfe0a9dab38f4423771c8ebced2020b1e15cbb35470c1ca2-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_encoding_format_base64[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375207,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-bbfbcf20cac146e0ae5e45ae6a42632d",
+              "object": "model_permission",
+              "created": 1762375207,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-7ed97509ff199eabe1380caa36b9e5934e9d04a9cafcfa2d21d20f6f85679ae4-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375273,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-4935d35e00fd4acdbe78662f42342e77",
+              "object": "model_permission",
+              "created": 1762375273,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-805e6b510b1ab33505a1af85c0d2a766cd3415512212d80f6292ca0ef5c359e1-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_base64_batch_processing[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374591,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-e19031997a1e44d99c8b5ae55725a887",
+              "object": "model_permission",
+              "created": 1762374591,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-860b7e8309e0761e20e845be75c0a28d759384a367f6308f2a921702318a5dba-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_non_streaming[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:non_streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375027,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-ec44b40a73b04912a837001376b59cff",
+              "object": "model_permission",
+              "created": 1762375027,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-8903569d538f9836ac6251d90c4668d3057e8e0ced847a08fd7a6faedb5710c3-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_single_string[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374356,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-3203232f1dbd426aba98ef1593dd3c01",
+              "object": "model_permission",
+              "created": 1762374356,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-8aa8c593dd64639678c294146fd56804393856c6e85197e6317ebd88351be21d-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming_with_n[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375248,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-5efe67a621074e979edaaf8fcfee9a80",
+              "object": "model_permission",
+              "created": 1762375248,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-8fc4c7b563b9bd423b74dcb4683039248f41d86c02703bd2dce845d972c9ae6f-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_user_parameter[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375135,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-5509cf924e5e4fc89091e4593f264258",
+              "object": "model_permission",
+              "created": 1762375135,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-99ae704b53e3e3150cac5cd579e446e6545a4ab6a63048ce00ee1fbe5fbf1b4e-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_non_streaming[openai_client-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:non_streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374301,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-72ed55b56df1471b9f71c48bacf8b768",
+              "object": "model_permission",
+              "created": 1762374301,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-9a36a281899f0800f085473f5f0185b09a02022607b965ac08b4db2e9e7eabc9-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_non_streaming_suffix[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:suffix]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374295,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-abbbfbb49abc4312b2b2011d4d2ba19b",
+              "object": "model_permission",
+              "created": 1762374295,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-9beac41c66cbe8568bb72b5ba0f5608597ef8a14b42585c22b1e7c45526537c1-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_single_string[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375065,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-d943fbda14264715906334300853cec7",
+              "object": "model_permission",
+              "created": 1762375065,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-a495ae010d48bb3649c822e3299e819c164c2311db231c81296ff4c72e6f81cb-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_non_streaming_with_file[txt=vllm/Qwen/Qwen3-0.6B]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374323,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-15a0a1106fff4fdd8ce7574373fe3cee",
+              "object": "model_permission",
+              "created": 1762374323,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-a77b3cb7370fd9f46e6ea12d72e1d9a8e7515f745289e93e5eb4a21d0e7b71b7-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_multiple_strings[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375082,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-ef1d3bc6fefc432380ef0eabdf216fd3",
+              "object": "model_permission",
+              "created": 1762375082,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-a82e913e058618dcb30b269a54d4e6a9cb1e0017a42efe04480874fe957194d4-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_invalid_model_error[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375165,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-b29f7386725b4f13976cd76b6dc3a278",
+              "object": "model_permission",
+              "created": 1762375165,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-bee98cb55c3b74854d0bb71b23b7e01bbb9f1580b413a26dc3afbf9da8b7d995-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_logprobs[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:log_probs]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374336,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-21db8cc1a31e41eaaa4e653435618645",
+              "object": "model_permission",
+              "created": 1762374336,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-c3d9f0302c09cecba4c3797ec2d65e358910e6194e13d1001fd3567ab2eff6aa-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_different_inputs_different_outputs[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374547,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-09fed2c5660e42658ab23c6d17b7840c",
+              "object": "model_permission",
+              "created": 1762374547,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-c6e251660301fe3f503b4c31dcb551087ca9118e65b97bd894954847723a9be0-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_completion_streaming[txt=vllm/Qwen/Qwen3-0.6B-inference:completion:sanity]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374297,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-70d68a901d2445f6b7f470c600b34c78",
+              "object": "model_permission",
+              "created": 1762374297,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-cb1f7d5cd412fddb3395ef125bbcdac95c85585f23684e71abf142004b164bbc-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store[openai_client-txt=vllm/Qwen/Qwen3-0.6B-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375047,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-a8446fd6718649399402526dc6fe1477",
+              "object": "model_permission",
+              "created": 1762375047,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-cbecbec285766025f2bebca94904e63578190f33b47eb6f32cb4635a1b43e3cf-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375254,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-f6eb51901e6443e492061deac904737c",
+              "object": "model_permission",
+              "created": 1762375254,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-d650458718dae3a10405ce1d241f0e1ceeeae8edf516cf10c611edcdf64035e3-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming_with_n[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_02]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375279,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-31e50ba39ad84a7daa1a24a3c77dc550",
+              "object": "model_permission",
+              "created": 1762375279,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-d8acc76e3d1b54eac9754a9d3a72c571fe3078b227a257aa15afdba946b69665-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_openai_chat_completion_streaming[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-inference:chat_completion:streaming_01]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375241,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-3ebcc379347541ea94de0f91838829e5",
+              "object": "model_permission",
+              "created": 1762375241,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-d9ff5f5ffaa7a64101936007fbe61cf2ed54f67609b54b56d92cb949234e3799-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_multiple_strings[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374449,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-311f880045284a469a286b8039177d10",
+              "object": "model_permission",
+              "created": 1762374449,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-da380037dc0fe8ae61b838baf268e616057e46f8424df0a9b52f94e48cef4a7f-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store_tool_calls[openai_client-txt=vllm/Qwen/Qwen3-0.6B-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375053,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-2e52800baf7e4d3389892f33feb3f52b",
+              "object": "model_permission",
+              "created": 1762375053,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-e42ca9261e3cee9c877322a51791ab6f113478170f8a21cd0a971c53b330e999-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_empty_list_error[llama_stack_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375150,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-cfec81fed838407597a92838017f3ef5",
+              "object": "model_permission",
+              "created": 1762375150,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-e5255919e39635597ad57c723896f9d258abaad9908b22ccd03c126ce597a5db-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_encoding_format_float[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374466,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-8811d359d9724f8cac7fd6df608f69bd",
+              "object": "model_permission",
+              "created": 1762374466,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-e6664ff0c07b13aa2af6a85925f3841eef3907bc4a55f8bc352a8c960e782ada-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_with_dimensions[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374482,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-d4b4923adfdf40b7bd7698aa798e68eb",
+              "object": "model_permission",
+              "created": 1762374482,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-edbd3344609a0fa1e97f75ede14a094a34db0dd6cb52975abae9f6e7832c6760-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store_tool_calls[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375291,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-a48cfd65bcd847d7aea01d44e8add51e",
+              "object": "model_permission",
+              "created": 1762375291,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-f6a9f5d7181cf078717443564e4de54e08845224d96b9c8150fb5cfda2068e82-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_embeddings.py::test_openai_embeddings_empty_list_error[openai_client-emb=sentence-transformers/nomic-ai/nomic-embed-text-v1.5]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762374517,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-916d53706b624fefb83e5dcc699e7a69",
+              "object": "model_permission",
+              "created": 1762374517,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/inference/recordings/models-f936269fe152d95db3fb80fb10482e3cc79cfd6a28ebdf1a7a8b220ba2de641b-fb68f5a6.json

@@ -0,0 +1,45 @@
+{
+  "test_id": "tests/integration/inference/test_openai_completion.py::test_inference_store[client_with_models-txt=vllm/Qwen/Qwen3-0.6B-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:8000/v1/v1/models",
+    "headers": {},
+    "body": {},
+    "endpoint": "/v1/models",
+    "model": ""
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.model.Model",
+        "__data__": {
+          "id": "Qwen/Qwen3-0.6B",
+          "created": 1762375285,
+          "object": "model",
+          "owned_by": "vllm",
+          "root": "/root/.cache/Qwen3-0.6B",
+          "parent": null,
+          "max_model_len": 8192,
+          "permission": [
+            {
+              "id": "modelperm-e0640be42b814b3394545ebe92d844b3",
+              "object": "model_permission",
+              "created": 1762375285,
+              "allow_create_engine": false,
+              "allow_sampling": true,
+              "allow_logprobs": true,
+              "allow_search_indices": false,
+              "allow_view": true,
+              "allow_fine_tuning": false,
+              "organization": "*",
+              "group": null,
+              "is_blocking": false
+            }
+          ]
+        }
+      }
+    ],
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/suites.py

@@ -78,7 +78,7 @@ SETUP_DEFINITIONS: dict[str, Setup] = {
             "VLLM_URL": "http://localhost:8000/v1",
         },
         defaults={
-            "text_model": "vllm/meta-llama/Llama-3.2-1B-Instruct",
+            "text_model": "vllm/Qwen/Qwen3-0.6B",
             "embedding_model": "sentence-transformers/nomic-embed-text-v1.5",
         },
     ),
@@ -169,6 +169,11 @@ SUITE_DEFINITIONS: dict[str, Suite] = {
         roots=base_roots,
         default_setup="ollama",
     ),
+    "base-vllm-subset": Suite(
+        name="base-vllm-subset",
+        roots=["tests/integration/inference"],
+        default_setup="vllm",
+    ),
     "responses": Suite(
         name="responses",
         roots=["tests/integration/responses"],

tests/integration/vector_io/test_openai_vector_stores.py

@@ -350,7 +350,7 @@ def test_openai_vector_store_search_empty(
     assert search_response is not None
     assert hasattr(search_response, "data")
     assert len(search_response.data) == 0  # Empty store should return no results
-    assert search_response.search_query == "test query"
+    assert search_response.search_query == ["test query"]
     assert search_response.has_more is False
 
 
@@ -679,7 +679,7 @@ def test_openai_vector_store_attach_file(
     assert file_attach_response.id == file.id
     assert file_attach_response.vector_store_id == vector_store.id
     assert file_attach_response.status == "completed"
-    assert file_attach_response.chunking_strategy.type == "auto"
+    assert file_attach_response.chunking_strategy.type == "static"
     assert file_attach_response.created_at > 0
     assert not file_attach_response.last_error
 
@@ -815,8 +815,8 @@ def test_openai_vector_store_list_files(
     assert set(file_ids) == {file.id for file in files_list.data}
     assert files_list.data[0].object == "vector_store.file"
     assert files_list.data[0].vector_store_id == vector_store.id
-    assert files_list.data[0].status == "completed"
+    assert files_list.data[0].status in ["completed", "in_progress"]
-    assert files_list.data[0].chunking_strategy.type == "auto"
+    assert files_list.data[0].chunking_strategy.type == "static"
     assert files_list.data[0].created_at > 0
     assert files_list.first_id == files_list.data[0].id
     assert not files_list.data[0].last_error
@@ -825,7 +825,7 @@ def test_openai_vector_store_list_files(
     assert first_page.has_more
     assert len(first_page.data) == 2
     assert first_page.first_id == first_page.data[0].id
-    assert first_page.last_id != first_page.data[-1].id
+    assert first_page.last_id == first_page.data[-1].id
 
     next_page = compat_client.vector_stores.files.list(
         vector_store_id=vector_store.id, limit=2, after=first_page.data[-1].id