Merge branch 'main' into prompt-api

llama_stack/providers/remote/inference/anthropic/anthropic.py

@@ -5,12 +5,13 @@
 # the root directory of this source tree.
 
 from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+from llama_stack.providers.utils.inference.openai_mixin import OpenAIMixin
 
 from .config import AnthropicConfig
 from .models import MODEL_ENTRIES
 
 
-class AnthropicInferenceAdapter(LiteLLMOpenAIMixin):
+class AnthropicInferenceAdapter(OpenAIMixin, LiteLLMOpenAIMixin):
     def __init__(self, config: AnthropicConfig) -> None:
         LiteLLMOpenAIMixin.__init__(
             self,
@@ -26,3 +27,8 @@ class AnthropicInferenceAdapter(LiteLLMOpenAIMixin):
 
     async def shutdown(self) -> None:
         await super().shutdown()
+
+    get_api_key = LiteLLMOpenAIMixin.get_api_key
+
+    def get_base_url(self):
+        return "https://api.anthropic.com/v1"

tests/integration/inference/test_openai_completion.py

@@ -41,6 +41,7 @@ def skip_if_model_doesnt_support_openai_completion(client_with_models, model_id)
         # or see the docs at https://console.groq.com/docs/","type":"invalid_request_error","code":"unknown_url"}}
         "remote::groq",
         "remote::gemini",  # https://generativelanguage.googleapis.com/v1beta/openai/completions -> 404
+        "remote::anthropic",  # at least claude-3-{5,7}-{haiku,sonnet}-* / claude-{sonnet,opus}-4-* are not supported
     ):
         pytest.skip(f"Model {model_id} hosted by {provider.provider_type} doesn't support OpenAI completions.")
 
@@ -73,6 +74,8 @@ def skip_if_doesnt_support_n(client_with_models, model_id):
         # Error code: 400 - [{'error': {'code': 400, 'message': 'Only one candidate can be specified in the
         # current model', 'status': 'INVALID_ARGUMENT'}}]
         "remote::gemini",
+        # https://docs.anthropic.com/en/api/openai-sdk#simple-fields
+        "remote::anthropic",
     ):
         pytest.skip(f"Model {model_id} hosted by {provider.provider_type} doesn't support n param.")