8b5b343841
attempt fix flaky tests
2026-01-23 12:10:08 -08:00
ca8c2c3938
fix #19620 : SSO user roles are not updated for existing users ( #19621 )
...
* Fix: SSO user roles are not updated for existing users
Fixes #19620
* Refactor: Remove redundant user_info retrieval in SSOAuthenticationHandler
* Test: add new tests for user creation and updates in get_user_info_from_db
2026-01-23 09:05:29 -08:00
9894721285
Merge pull request #19548 from BerriAI/litellm_staging_01_22_2026
...
Litellm staging 01 22 2026
2026-01-23 20:03:11 +05:30
8ac1d96d90
Merge pull request #19634 from BerriAI/litellm_feat_hashicorp_rotate
...
[feat] hashicorp vault rotate support
2026-01-23 21:08:55 +09:00
12463809bd
Merge pull request #19638 from BerriAI/main
...
merge main in stagin 1 22 26
2026-01-23 14:54:17 +05:30
695fbf4ec5
feat: hashicorp vault rotate support
2026-01-23 17:32:55 +09:00
919033a6d0
fix: include tool arguments in proxy_server_request for spend logs callbacks
2026-01-23 16:36:37 +09:00
12bc66aa5b
Merge pull request #19623 from BerriAI/litellm_fix_completions_mcp_output_ordering
...
[fix] completions mcp output ordering
2026-01-23 15:56:02 +09:00
6a60b3d848
test: completions mcp output test
2026-01-23 15:17:14 +09:00
26a2c90818
[Fix] Anthropic models on Azure AI cache pricing ( #19532 ) ( #19614 )
2026-01-22 20:00:40 -08:00
69c8698e62
fix: pass through endpoints update registry ( #19420 )
...
* fix: pass through endpoints update registry
* add test case, fix lint error and comment to avoid confusion
* fix pass through endpoints test case
2026-01-22 19:57:48 -08:00
c23e4b87dc
[Feat] New LiteLLM Policy engine - create policies to manage guardrails, conditions - permissions per Key, Team ( #19612 )
...
* init PolicyMatcher
* TestPolicyMatcherGetMatchingPolicies
* TestPolicyMatcherGetMatchingPolicies
* feat: init PolicyResolver
* init resolver types
* init policy from config
* inint PolicyValidator
* validate policy
* init Architecture Diagram
* test_add_guardrails_from_policy_engine
* init _init_policy_engine
* test updates
* test fixws
* new attachment config
* simplify types
* TestPolicyResolverInheritance
* fix policy resolver
* fix policies
* fix applied policy
* docs fix
* docs fix
* fix linting + QA checks
* fix linting + QA fixes
* test fixes
2026-01-22 19:49:53 -08:00
88f8f49e1d
fix(websearch_interception): filter internal kwargs before follow-up request ( #19577 )
...
The websearch interception handler was passing internal flags like
`_websearch_interception_converted_stream` to the follow-up LLM request.
This caused "Extra inputs are not permitted" errors from providers like
Bedrock that use strict Pydantic validation.
Fix: Filter out all kwargs starting with `_websearch_interception` prefix
before making the follow-up anthropic_messages.acreate() call.
2026-01-22 10:42:20 -08:00
a51835dfcc
Metrics prometheus user team count ( #19520 )
...
* add user count and team count prometheus metrics
* rebase
* revert mistaken deletion
2026-01-22 08:17:15 -08:00
842e5b3ad6
Merge pull request #19560 from BerriAI/litellm_bedrock_invoke_structured_output
...
[Feat] Add support for output formatfor bedrock invoke via v1/messages
2026-01-22 19:44:48 +05:30
c78c878822
Merge pull request #19558 from BerriAI/litellm_gemini_vertexai_mapping
...
Add custom vertex ai mapping to the output
2026-01-22 19:44:32 +05:30
73715ab417
Merge pull request #19556 from BerriAI/litellm_fix_gemini_batch_jan22
...
Fix: generation config empty for batch
2026-01-22 19:44:26 +05:30
f312bf23d0
Fix:test_multiple_function_call
2026-01-22 19:34:54 +05:30
991fee056f
Fix batch tests
2026-01-22 19:23:32 +05:30
8a622c51f5
feat: Add MCP tools response to chat completions
2026-01-22 19:23:32 +05:30
ad1edd38d5
Merge branch 'main' into litellm_staging_01_21_2026
2026-01-22 17:56:40 +05:30
24faca9bcf
Add support for output formatfor bedrock invoke via v1/messages
2026-01-22 16:36:03 +05:30
18240662db
Add custom vertex ai mapping to the output
2026-01-22 15:18:24 +05:30
ed67bf2705
feat: Add MCP tools response to chat completions
2026-01-22 15:32:04 +09:00
9f57eb3e74
Fix Azure AI costs for Anthropic models ( #19530 )
...
* Fix Azure AI cost calculation
* fixup
2026-01-21 21:10:27 -08:00
a3f7f5858b
Fix date overflow/division by zero in proxy utils ( #19527 )
...
* Fix date overflow/division by zero in proxy utils
* Fix projected spend calculation
* Strengthen projected spend tests
2026-01-21 21:09:57 -08:00
ab274ac3c4
fix(azure response api): flatten tools for responses api to support nested definitions ( #19526 )
...
The Azure Responses API uses a different schema (flattened) for tools compared to the standard OpenAI/Azure Chat Completions API (nested). This caused a `BadRequestError` when users passed standard tool definitions.
Changes:
- Implemented tool flattening logic in `AzureOpenAIResponsesAPIConfig.transform_responses_api_request`.
- Added comprehensive unit tests in test_azure_transformation.py to verify nested-to-flat transformation, pass-through of flat tools, and immutability.
- Ensures cross-provider compatibility for tool definitions.
Fixes #19523
2026-01-21 21:08:28 -08:00
60840ea292
fix(bedrock): correct streaming choice index for tool calls ( #19506 )
...
Bedrock's contentBlockIndex identifies content blocks within a message
(text=0, tool_call=1), not OpenAI's choice index (which varies with n>1).
This caused OpenAI SDK's ChatCompletionAccumulator to fail when tool call
chunks arrived on index 1 while finish_reason arrived on index 0.
Bedrock doesn't support n>1 (no such parameter exists):
https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InferenceConfiguration.html
OpenAI choice index spec:
https://platform.openai.com/docs/api-reference/chat/streaming
2026-01-21 20:57:14 -08:00
746414eb9b
Fix/per service ssl override v2 ( #19538 )
...
* refactor(ssl): support per-service SSL verification overrides
* add test cases for ssl
2026-01-21 20:10:04 -08:00
7777aeb695
fixing prompt-security's guardrail implementation ( #19374 )
...
* Consolidated change
* fix(prompt_security): update message processing to persist sanitized files and filter for API calls
* fix per krrishdholakia suggestion
2026-01-21 20:09:40 -08:00
363b0cc132
fix(azure): preserve content_policy_violation details for images ( #19328 ) ( #19372 )
...
Azure OpenAI Images (DALL·E 3) returns policy violations as a structured payload under body["error"], including inner_error.content_filter_results and revised_prompt.
LiteLLM previously:
- Failed to extract nested error messages (get_error_message only handled body["message"])
- Missed policy violation detection when error strings were generic
- Dropped inner_error details when raising ContentPolicyViolationError
This change:
- Extracts nested Azure error fields (code/type/message + inner_error)
- Detects policy violations via structured error codes
- Passes an OpenAI-style error body + provider_specific_fields to preserve details
Tests:
- python3 -m pytest tests/test_litellm/llms/azure/test_azure_exception_mapping.py
- python3 -m pytest tests/test_litellm/litellm_core_utils/test_exception_mapping_utils.py
Fixes #19328
2026-01-21 20:06:51 -08:00
0e738a5027
fix(mcp): forward static_headers to MCP servers ( #19341 ) ( #19366 )
...
Forward static_headers from /mcp-rest/test/* routes into the MCP client so headers are present during session.initialize() and tool discovery.
Also add a shared merge_mcp_headers() helper to keep header precedence consistent and ensure OpenAPI-to-MCP generated tools include static_headers.
Tests:
- pytest tests/test_litellm/proxy/_experimental/mcp_server/test_rest_endpoints.py
- pytest tests/test_litellm/proxy/_experimental/mcp_server/test_mcp_server_manager.py -k register_openapi_tools_includes_static_headers
Fixes #19341
Co-authored-by: Krish Dholakia <krrishdholakia@gmail.com >
2026-01-21 19:30:55 -08:00
d5e912322f
[Fix] VertexAI Pass through - Ensure only anthropic betas are forwarded down to LLM API ( #19542 )
...
* fix ALLOWED_VERTEX_AI_PASSTHROUGH_HEADERS
* test_vertex_passthrough_forwards_anthropic_beta_header
* fix test_vertex_passthrough_forwards_anthropic_beta_header
* test_vertex_passthrough_does_not_forward_litellm_auth_token
* fix utils
* Using Anthropic Beta Features on Vertex AI
* test_forward_headers_from_request_x_pass_prefix
2026-01-21 19:12:04 -08:00
6b6785bc4f
Merge pull request #19539 from BerriAI/litellm_models_scope
...
[Feature] Adding Optional scope Param to /models
2026-01-21 17:41:22 -08:00
c6b157832b
Merge pull request #19296 from BerriAI/litellm_esca_reissue
...
[Reissue: Fix] /user/new Privilege Escalation
2026-01-21 16:46:34 -08:00
6723b30d03
Adding scope to /models
2026-01-21 16:40:31 -08:00
4a14a53ae8
Merge pull request #19469 from BerriAI/litellm_feat_mcp_spendlogs
...
[feat] mcp spendlogs
2026-01-22 05:29:21 +09:00
5cb5969a26
[Fix] LiteLLM VertexAI Pass through - ensuring incoming headers are forwarded down to target ( #19524 )
...
* test_vertex_passthrough_forwards_anthropic_beta_header
* add_incoming_headers
2026-01-21 12:01:33 -08:00
d0e35751a1
Fixing tests and linting
2026-01-21 11:02:39 -08:00
b5a7d2ab34
Paginating model/info endpoint
2026-01-21 10:44:18 -08:00
aa4b0e0149
Fix duplicate test_handler.py filenames causing pytest collection errors ( #19385 )
2026-01-21 08:47:50 -08:00
e758dd0a59
Merge pull request #19472 from BerriAI/litellm_fix_chat_completion_responses_streaming
...
Fix: tool call streaming in chat completion bridge
2026-01-21 19:15:53 +05:30
b6691baf24
Merge pull request #19482 from BerriAI/litellm_anthropic_max_length
...
Filter out unsupported fields from JSON schema for Anthropic's output_format API.
2026-01-21 18:57:16 +05:30
d879dcdcef
Revert "Fix/prisma schema permission ( #19391 )"
...
This reverts commit 75ee0d126c
2026-01-21 17:19:32 +05:30
a1aba2ed8d
Merge pull request #19491 from BerriAI/main
...
merge main 20 1 25
2026-01-21 16:40:48 +05:30
c571a02f3b
Filter out unsupported fields from JSON schema for Anthropic's output_format API.
2026-01-21 15:48:22 +05:30
3a0d166eb0
Fix: tool call streaming in chat completino brigde
2026-01-21 12:14:02 +05:30
caf5f7f8ae
test: add test
2026-01-21 14:51:56 +09:00
15013cec4b
feat(gemini): add file content support in tool results ( #19416 )
...
Add support for 'file' and 'input_file' content types in
convert_to_gemini_tool_call_result(). File content in tool
results was previously silently dropped.
Supports base64 data URIs and HTTP URLs, matching the existing
image handling pattern. Enables PDF, audio, video, and other
file types as inline_data for Gemini.
2026-01-20 19:54:12 -08:00
b4ed387d24
fix(vertex_ai): handle reasoning_effort as dict from OpenAI Agents SDK ( #19419 )
...
The OpenAI Agents SDK (v0.6.9+) now passes reasoning_effort as a dict
when summary is specified: {"effort": "high", "summary": "auto"}
This change extracts the "effort" value from the dict for Vertex AI,
which only supports thinkingLevel (not summary).
Before: reasoning_effort={"effort": "high"} was silently ignored
After: reasoning_effort={"effort": "high"} correctly maps to thinkingLevel
Fixes #19411
2026-01-20 19:31:25 -08:00
yuneng-jiang