Add x-api-key header to CountTokens handler to match chat completion
authentication. Azure AI Anthropic requires this header per Microsoft's
native API format.
Bedrock rejects requests when toolResult or toolUse blocks within a
single message contain duplicate IDs. The Converse message transformer
merges consecutive tool/assistant messages without checking for
duplicate toolUseId values, causing BedrockException errors.
Add _deduplicate_bedrock_content_blocks() — a generalized helper that
removes duplicate blocks by ID, logs a warning for each dropped
duplicate via verbose_logger, and preserves non-tool blocks (e.g.
cachePoint). Apply it at all four merge sites (sync/async × toolResult/
toolUse).
The Anthropic /messages path was fixed in PR #19324; this applies the
equivalent fix to the Bedrock Converse path.
Fixes#20048
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
* Fix Nova grounding web_search_options={} not applying systemTool
Two bugs prevented web_search_options={} from working for Nova grounding:
1. Empty dict falsy check: The condition `value and isinstance(value, dict)`
short-circuits to False when value is {} (empty dict is falsy in Python).
Changed to `isinstance(value, dict)` to match Anthropic's implementation.
2. Pre-formatted tools mangled by _bedrock_tools_pt: The systemTool
(already in Bedrock format) was added to optional_params["tools"], but
_process_tools_and_beta passed all tools through _bedrock_tools_pt which
expects OpenAI-format tools. This corrupted the systemTool into an empty
toolSpec. Fixed by separating systemTool blocks before transformation
and appending them after.
Fixes follow-up to #19598
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Fix python-multipart Python version constraint for Poetry lock
python-multipart ^0.0.22 requires Python >=3.10 but the project supports
>=3.9. Add python = ">=3.10" marker so Poetry can resolve dependencies
for Python 3.9.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Previously, when Model Armor guardrail blocked a request/response,
the `applied_guardrails` field was not populated in the logs because
`add_guardrail_to_applied_guardrails_header()` was called after the
HTTPException was raised.
This fix moves the `add_guardrail_to_applied_guardrails_header()` call
to before the blocking check in all hooks:
- async_pre_call_hook (pre_call mode)
- async_moderation_hook (during_call mode)
- async_post_call_success_hook (post_call mode)
- async_post_call_streaming_iterator_hook (streaming)
This ensures that even when a guardrail blocks content, the guardrail
name is properly recorded in the logs for observability.
Added regression tests to verify applied_guardrails is populated when
content is blocked.
Co-authored-by: Cursor <cursoragent@cursor.com>
When using LiteLLM's Anthropic /v1/messages endpoint to route requests to
OpenAI models, requests fail if any tool name exceeds OpenAI's 64-character
limit. Anthropic API has no such limit, causing compatibility issues.
Changes:
- Add truncate_tool_name() function using {55-char-prefix}_{8-char-hash} format
- Modify translate_anthropic_tools_to_openai() to truncate and return mapping
- Modify translate_anthropic_tool_choice_to_openai() to truncate tool name
- Restore original tool names in responses using the mapping
- Support tool name restoration in streaming responses
- Add backwards-compatible API (existing methods still work)
The fix only applies when routing Anthropic requests to OpenAI models.
Native Anthropic/Claude requests pass through unchanged.
* fix: models loadbalancing billing issue by filter (#18891)
* fix: models loadbalancing billing issue by filter
* fix: separate key and team access groups in metadata
* fix: lint issues
* Add async_post_call_response_headers_hook to CustomLogger (#20070)
Allow CustomLogger callbacks to inject custom HTTP response headers
into streaming, non-streaming, and failure responses via a new
async_post_call_response_headers_hook method.
* async_post_call_response_headers_hook
---------
Co-authored-by: michelligabriele <gabriele.michelli@icloud.com>
* Add /realtime API benchmarks to Benchmarks documentation
- Added new section showing performance improvements for /realtime endpoint
- Included before/after metrics showing 182× faster p99 latency
- Added test setup specifications and key optimizations
- Referenced from v1.80.5-stable release notes
Co-authored-by: ishaan <ishaan@berri.ai>
* Update /realtime benchmarks to show current performance only
- Removed before/after comparison, showing only current metrics
- Clarified that benchmarks are e2e latency against fake realtime endpoint
- Simplified table format for better readability
Co-authored-by: ishaan <ishaan@berri.ai>
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: ishaan <ishaan@berri.ai>
* Add LangSmith mock client support
- Create langsmith_mock_client.py following GCS and Langfuse patterns
- Add mock mode detection via LANGSMITH_MOCK environment variable
- Intercept LangSmith API calls via AsyncHTTPHandler.post patching
- Add verbose logging throughout mock implementation
- Update LangsmithLogger to initialize mock client when mock mode enabled
- Supports configurable mock latency via LANGSMITH_MOCK_LATENCY_MS
* Add Datadog mock client support
- Create datadog_mock_client.py following GCS, Langfuse, and LangSmith patterns
- Add mock mode detection via DATADOG_MOCK environment variable
- Intercept Datadog API calls via AsyncHTTPHandler.post and httpx.Client.post patching
- Add verbose logging throughout mock implementation
- Update DataDogLogger and DataDogLLMObsLogger to initialize mock client when mock mode enabled
- Supports both async and sync logging paths
- Supports configurable mock latency via DATADOG_MOCK_LATENCY_MS
* refactor: consolidate mock client logic into factory pattern
- Create mock_client_factory.py to centralize common mock HTTP client logic
- Refactor GCS, Langfuse, LangSmith, and Datadog mock clients to use factory
- Improve GET/DELETE mock accuracy for GCS (return valid StandardLoggingPayload)
- Fix DELETE mock to return empty body (204 No Content) instead of JSON
- Reduce code duplication across integration mock clients
* feat: add PostHog mock client support
- Create posthog_mock_client.py using factory pattern
- Integrate mock client into PostHogLogger with mock mode detection
- Add verbose logging for mock mode initialization and batch operations
- Enable mock mode via POSTHOG_MOCK environment variable
* Add Helicone mock client support
- Created helicone_mock_client.py using factory pattern (similar to GCS)
- Integrated mock mode detection and initialization in HeliconeLogger
- Mock client patches HTTPHandler.post to intercept Helicone API calls
- Uses factory pattern for should_use_mock and MockResponse utilities
- Custom HTTPHandler.post patching required since HTTPHandler uses self.client.send()
* Add mock support for Braintrust integration and extend mock client factory
- Add braintrust_mock_client.py with mock HTTP client for Braintrust integration testing
- Integrate mock client into BraintrustLogger with mock mode detection
- Refactor Helicone mock client to fully utilize factory's HTTPHandler.post patching
- Extend mock_client_factory to support patching HTTPHandler.post for sync calls
- Enable endpoint-specific mock responses for Braintrust (/project vs /project_logs)
- All mock clients now properly handle both async (AsyncHTTPHandler) and sync (HTTPHandler) calls
* Fix linter errors: remove unused imports and suppress complexity warning
- Remove unused imports from gcs_bucket_mock_client.py (httpx, json, timedelta, Dict, Optional)
- Remove unused Callable import from mock_client_factory.py
- Add noqa comment to suppress PLR0915 complexity warning for create_mock_client_factory function
* Document mock environment variables for PostHog, Helicone, Braintrust, Datadog, and Langsmith integrations
- Add POSTHOG_MOCK and POSTHOG_MOCK_LATENCY_MS documentation
- Add HELICONE_MOCK and HELICONE_MOCK_LATENCY_MS documentation
- Add BRAINTRUST_MOCK and BRAINTRUST_MOCK_LATENCY_MS documentation
- Add DATADOG_MOCK and DATADOG_MOCK_LATENCY_MS documentation
- Add LANGSMITH_MOCK and LANGSMITH_MOCK_LATENCY_MS documentation
All mock env vars follow the same pattern: enable mock mode for integration testing by intercepting API calls and returning mock responses without making actual network calls.
* Fix security issue
