Commit graph

32501 commits

Author SHA1 Message Date
Yuneng Jiang
ba0c0864ca
chore: fixes 2026-04-04 23:47:13 -07:00
Ishaan Jaffer
c0f13c7153 test: Add unit test for migration failure exit behavior 2026-02-16 10:23:52 -08:00
Ishaan Jaffer
15661d15c4 fix: Exit proxy if database migrations fail 2026-02-16 09:33:34 -08:00
Krish Dholakia
45690db820
UI - Content Filters, help edit/view categories and 1-click add categories + go to next page (#21223)
* feat(ui/): allow viewing content filter categories on guardrail info

* fix(add_guardrail_form.tsx): add validation check to prevent adding empty content filter guardrails

* feat(ui/): improve ux around adding new content filter categories

easy to skip adding a category, so make it a 1-click thing
2026-02-16 09:11:16 -08:00
michelligabriele
035f0916ad
fix(mcp): revert StreamableHTTPSessionManager to stateless mode (#21323)
PR #19809 changed stateless=True to stateless=False to enable progress
notifications for MCP tool calls. This caused the mcp library to enforce
mcp-session-id headers on all non-initialize requests, breaking MCP
Inspector, curl, and any client without automatic session management.

Revert to stateless=True to restore compatibility with all MCP clients.
The progress notification code already handles missing sessions gracefully
(defensive checks + try/except), so no other changes are needed.

Fixes #20242
2026-02-16 09:08:44 -08:00
Vincent Koc
0dcc744f7e
fix(proxy): handle missing DATABASE_URL in append_query_params (#21239)
* fix: handle missing database url in append_query_params

* Update litellm/proxy/proxy_cli.py

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

---------

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
2026-02-16 09:03:10 -08:00
Krrish Dholakia
89e95f779f fix(index.md): cleanup str 2026-02-16 08:59:17 -08:00
Atharva Jaiswal
7bcef1490b
Fix: Exclude tool params for models without function calling support (#21125) (#21244)
* Fix tool params reported as supported for models without function calling (#21125)

JSON-configured providers (e.g. PublicAI) inherited all OpenAI params
including tools, tool_choice, function_call, and functions — even for
models that don't support function calling. This caused an inconsistency
where get_supported_openai_params included "tools" but
supports_function_calling returned False.

The fix checks supports_function_calling in the dynamic config's
get_supported_openai_params and removes tool-related params when the
model doesn't support it. Follows the same pattern used by OVHCloud
and Fireworks AI providers.

* Style: move verbose_logger to module-level import, remove redundant try/except

Address review feedback from Greptile bot:
- Move verbose_logger import to top-level (matches project convention)
- Remove redundant try/except around supports_function_calling() since it
  already handles exceptions internally via _supports_factory()
2026-02-16 08:36:32 -08:00
Sameer Kankute
351e834802
Merge pull request #21320 from BerriAI/litellm_anthropic_doc_beta_header
Litellm anthropic doc beta header
2026-02-16 22:03:18 +05:30
Sameer Kankute
452f481fa0 correct the time 2026-02-16 21:59:54 +05:30
Sameer Kankute
e67641cdb3 Add blog fffor Managing Anthropic Beta Headers 2026-02-16 21:46:12 +05:30
Sameer Kankute
2ad648a083 Add blog fffor Managing Anthropic Beta Headers 2026-02-16 21:45:01 +05:30
jquinter
bf93ce8841
Merge pull request #21284 from BerriAI/fix/mcp-server-test-mocks
fix(test): add cleanup fixture and no_parallel mark for MCP tests
2026-02-16 12:12:58 -03:00
jquinter
c3fb5e1ea5
Update tests/test_litellm/proxy/_experimental/mcp_server/test_mcp_server.py
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
2026-02-16 12:12:48 -03:00
jquinter
d84336bd07
Merge pull request #21283 from BerriAI/fix/video-generation-logging-mock
fix(test): correct async mock for video generation logging test
2026-02-16 12:12:20 -03:00
jquinter
87861549d8
Merge pull request #21282 from BerriAI/fix/cost-calculation-log-test-parallel
fix(test): replace caplog with custom handler for parallel execution
2026-02-16 12:11:48 -03:00
jquinter
a2ce7fc99d
Merge pull request #21280 from BerriAI/fix/add-pytest-postgresql-dependency
fix(deps): add pytest-postgresql for db schema migration tests
2026-02-16 12:09:55 -03:00
Julio Quinteros Pro
af9b6f6e0d Merge main into fix/add-pytest-postgresql-dependency
Resolved poetry.lock conflict by regenerating with Poetry 2.3.2.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-16 12:03:20 -03:00
Harshit Jain
1308a2cd76
Merge pull request #20314 from Harshit28j/fix/sso_PKCE_deployments
fix: SSO PKCE support fails in multi-pod Kubernetes deployments
2026-02-16 19:40:44 +05:30
Harshit Jain
dd2111e664
Merge pull request #20321 from Harshit28j/litellm_fix-virtual-key-grace-period
fix: virutal key grace period from env/UI
2026-02-16 19:36:18 +05:30
Sameer Kankute
86a254a215
Merge pull request #21307 from BerriAI/litellm_oss_staging_02_14_20262
Litellm oss staging 02 14 20262
2026-02-16 19:10:59 +05:30
Sameer Kankute
0f2b0869ae
Merge pull request #21311 from BerriAI/litellm_add_agents_sdk_openai
Add doc for OpenAI Agents SDK with LiteLLM
2026-02-16 19:06:27 +05:30
Sameer Kankute
6ceebdcfe2 fix mypy 2026-02-16 19:05:06 +05:30
Sameer Kankute
be9df253dd
Update docs/my-website/sidebars.js
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
2026-02-16 19:03:15 +05:30
Sameer Kankute
5b6e232da6 Add doc for OpenAI Agents SDK with LiteLLM 2026-02-16 19:00:53 +05:30
Sameer Kankute
a38b4c8924 Add doc for OpenAI Agents SDK with LiteLLM 2026-02-16 18:59:39 +05:30
Sameer Kankute
4548d9fbe9 fix mypy 2026-02-16 18:55:21 +05:30
Sameer Kankute
ac2f17a514 fix: proxy/batches_endpoints/endpoints.py:309:11: PLR0915 Too many statements (54 > 50) 2026-02-16 18:41:06 +05:30
Sameer Kankute
27890dd46e
Merge pull request #21306 from mjkam/fix/bedrock-min-budget-tokens
fix(bedrock): clamp thinking.budget_tokens to minimum 1024
2026-02-16 18:31:28 +05:30
Sameer Kankute
72a1bd66c7
Merge pull request #21157 from Point72/ephrimstanley/s3-logger-skip-missing-standard-logging-object
Managed batches - Misc bug fixes
2026-02-16 18:29:59 +05:30
Emerson Gomes
06e7bfce2e perf(spend): avoid duplicate daily agent transaction computation (#21187) 2026-02-16 18:28:34 +05:30
Constantine
7ef9083812 fix(aiohttp): prevent closing shared ClientSession in AiohttpTransport (#21117)
When a shared ClientSession is passed to LiteLLMAiohttpTransport,
calling aclose() on the transport would close the shared session,
breaking other clients still using it.

Add owns_session parameter (default True for backwards compatibility)
to AiohttpTransport and LiteLLMAiohttpTransport. When a shared session
is provided in http_handler.py, owns_session=False is set to prevent
the transport from closing a session it does not own.

This aligns AiohttpTransport with the ownership pattern already used
in AiohttpHandler (aiohttp_handler.py).
2026-02-16 18:28:28 +05:30
Kristoffer Arlind
51b1b0339c Allow effort="max" for Claude Opus 4.6 (#21112) 2026-02-16 18:28:22 +05:30
Ephrim Stanley
a3762e7d49 Addressed greptile comments to extract common helpers and return 404 2026-02-16 07:58:04 -05:00
Fly
9c71d8b61b feat(s3): add support for virtual-hosted-style URLs (#21094)
Add s3_use_virtual_hosted_style parameter to support AWS S3 virtual-hosted-style URL format (bucket.endpoint/key) alongside the existing path-style format (endpoint/bucket/key).

This enables compatibility with S3-compatible services like MinIO and aligns with AWS S3 official terminology.
2026-02-16 18:26:44 +05:30
jquinter
43f9a588d9 fix: improve Langfuse test isolation to prevent flaky failures (#21093)
The test was creating fresh mocks but not fully isolating from setUp state,
causing intermittent CI failures with 'Expected generation to be called once.
Called 0 times.'

Instead of creating fresh mocks, properly reset the existing setUp mocks to
ensure clean state while maintaining proper mock chain configuration.
2026-02-16 18:26:37 +05:30
Sameer Kankute
70a49a4b35
Merge pull request #21295 from BerriAI/litellm_correct_converse_usage
Fix converse anthropic usage object according to v1/messages specs
2026-02-16 18:22:30 +05:30
Sameer Kankute
4976029ee6
Merge pull request #21302 from BerriAI/litellm_kilo_ai
Add routing based on if reasoning is supported or not
2026-02-16 18:20:25 +05:30
mjkam
37da38fdaa fix(bedrock): clamp thinking.budget_tokens to minimum 1024
Bedrock rejects thinking.budget_tokens values below 1024 with a 400
error. This adds automatic clamping in the LiteLLM transformation
layer so callers (e.g. router with reasoning_effort="low") don't
need to know about the provider-specific minimum.

Fixes #21297

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-16 21:49:41 +09:00
Sameer Kankute
3347fabe6b
Merge pull request #21209 from jayy-77/fix/21193-chatgpt-codex-unsupported-params
Fix/21193 chatgpt codex unsupported params
2026-02-16 18:13:10 +05:30
Ephrim Stanley
cc4e022176 Removed stray .md file 2026-02-16 07:14:42 -05:00
Sameer Kankute
c00c5a6e29 add fireworks_ai/accounts/fireworks/models/kimi-k2p5 in model map 2026-02-16 17:39:04 +05:30
Sameer Kankute
ee9e997755 Add routing based on if reasoning is supported or not 2026-02-16 17:02:34 +05:30
Sameer Kankute
01cdec5771 Fix converse anthropic usage object according to v1/messages specs 2026-02-16 14:03:31 +05:30
Julio Quinteros Pro
f2b6c38c86 Remove redundant import inside test method
The module litellm.videos.main is already imported at the top of
the file (line 21), so the import inside the test method is redundant.

Addresses Greptile feedback (minor style issue).

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-15 21:15:52 -03:00
Julio Quinteros Pro
fe691cac08 Remove unused pytest import and add trailing newline
- Removed unused pytest import (caplog fixture was removed)
- Added missing trailing newline at end of file

Addresses Greptile feedback (minor style issues).

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-15 21:15:12 -03:00
Julio Quinteros Pro
1deb4456e0 Regenerate poetry.lock with Poetry 2.3.2
Updated lock file to use Poetry 2.3.2 (matching main branch standard).
This addresses Greptile feedback about Poetry version mismatch.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-15 21:12:22 -03:00
jquinter
44bb1dafdb
Merge pull request #21286 from BerriAI/fix/proxy-config-callback-validation
fix(test): mock environment variables for callback validation test
2026-02-15 21:11:12 -03:00
jquinter
49fd1b5de4
Merge pull request #21279 from BerriAI/fix/tokenizer-cache-test-isolation
fix(test): clear tokenizer LRU cache for test isolation
2026-02-15 21:05:38 -03:00
jquinter
5be76dc6e0
Merge pull request #21281 from BerriAI/fix/add-fakeredis-dependency
fix(deps): add fakeredis for pod lock manager tests
2026-02-15 21:04:54 -03:00