scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-17 23:51:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 66
35074 commits 2515 branches 1410 tags 3 GiB
Commit graph

35074 commits

This branch
This branch
All branches
Author SHA1 Message Date
Josh Bonczkowski
e3dac64fd1 Merge branch 'main' into newrelic 2026-03-11 16:39:25 -04:00
yuneng-jiang
b5dfde892c
Merge pull request #23377 from BerriAI/litellm_fix_flaky_mcp_and_agentcore_tests 
[Fix] Flaky MCP Server and AgentCore Streaming Tests
 2026-03-11 13:20:59 -07:00
Mr. Ånand
05fba27b0c
Add Retool Assist tutorial with LiteLLM Proxy to docs (#21952) 
* docs: add Retool Assist integration guide

- Add tutorials/retool_assist.md with setup instructions
- Add screenshots: Resources screen, Custom Provider config, resource query test
- Add retool_assist to AI Tools sidebar

Co-authored-by: Cursor <cursoragent@cursor.com>

* docs: refine Retool Assist guide layout

Co-authored-by: Cursor <cursoragent@cursor.com>

* docs: update Retool Assist guide with new screenshots

- Add Resources screen after step 1
- Update AI category and LiteLLM config modal images

Co-authored-by: Cursor <cursoragent@cursor.com>

* Apply suggestion from @greptile-apps[bot]

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

* updated blank line

* Added video & gifs

Made-with: Cursor

---------

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
 2026-03-11 13:17:08 -07:00
yuneng-jiang
64d3d7626f [Fix] Flaky MCP server and AgentCore streaming tests in CI 
- MCP tests: set mock_mcp_server.oauth2_flow = None to prevent MagicMock
  leaking into Pydantic Literal validation for MCPServer
- AgentCore tests: pass api_key="test-jwt-token" to bypass SigV4 credential
  lookup that fails in CI without AWS credentials

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-11 13:08:48 -07:00
yuneng-jiang
08bb3808bc
Merge pull request #23374 from BerriAI/ui_build_fix_mar11 
[infra] Fixing UI Build and Tests
 2026-03-11 12:55:58 -07:00
yuneng-jiang
d415ffc70b fixing ui build and tests 2026-03-11 12:55:16 -07:00
Josh Bonczkowski
c6a40767c5 Adding unit tests for the New Relic extension. 2026-03-11 15:53:06 -04:00
Cesar Garcia
59c24b7405
Merge pull request #17717 from Chesars/docs/responses-api-bridge-clarification 
docs: clarify when to use openai/responses/ prefix for built-in tools
 2026-03-11 15:18:32 -03:00
Cesar Garcia
57e73c8f5e
Merge pull request #20795 from Chesars/chore/remove-deprecated-models 
chore: cleanup deprecated models from pricing JSON
 2026-03-11 15:18:13 -03:00
Chesars
1a3fdc7ae3 fix: align Vertex AI Claude deprecations with Google's schedule 
- Restore vertex_ai/claude-3-7-sonnet@20250219 (Vertex AI shutdown is
  May 11, 2026, still active — was incorrectly removed based on
  Anthropic API retirement date)
- Remove vertex_ai/claude-3-5-sonnet-v2 and
  vertex_ai/claude-3-5-sonnet-v2@20241022 (Vertex AI shutdown was
  Feb 19, 2026, already past)
- Remove deprecated claude-3-7-sonnet-20250219 from web search test,
  use only non-deprecated models

Source: https://docs.google.com/vertex-ai/generative-ai/docs/deprecations/partner-models
 2026-03-11 15:11:07 -03:00
yuneng-jiang
1d1baa2cd5
Merge pull request #23362 from BerriAI/litellm_ui_model_mcp 
[Fix] Null safe checks for MCP Hub Tab
 2026-03-11 10:50:54 -07:00
Chesars
d81d751af0 fix(tests): update tests to use models still present in pricing JSON 
Replace removed deprecated models (claude-3-5-sonnet-20241022,
claude-3-5-haiku-20241022, claude-3-5-haiku-latest) with current
models in web_search and cost calculation tests.
 2026-03-11 14:50:47 -03:00
Josh Bonczkowski
e2be9959db Moving New Relic imports into the methods to support unit tests. 2026-03-11 13:37:25 -04:00
yuneng-jiang
8af34ffebc null-safe array checks 2026-03-11 10:35:51 -07:00
Chesars
45ca8b5b2b chore: resolve merge conflicts with upstream/main 
Accept upstream's new models and re-apply removal of 133 deprecated models.
 2026-03-11 14:24:53 -03:00
Chesars
f9a538b583 fix(docs): close unclosed code block before Examples heading 2026-03-11 14:15:54 -03:00
Cesar Garcia
274bf42493
Update docs/my-website/docs/providers/openai.md 
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
 2026-03-11 14:13:56 -03:00
Josh Bonczkowski
2236066c7f Merge branch 'main' into newrelic 2026-03-11 13:12:03 -04:00
Josh Bonczkowski
207692cec4 Updating LiteLLM documentation for New Relic extension. 2026-03-11 13:10:57 -04:00
Chesars
9e7a6a73ed docs: remove duplicate gpt-5.4 tip block 2026-03-11 13:48:47 -03:00
Chesars
01a6c707a3 docs: restore gpt-5.4 reasoning_effort tip lost during rebase 2026-03-11 13:48:07 -03:00
Chesars
d0d09e037e docs: clarify when to use openai/responses/ prefix for built-in tools 
The existing documentation for the Responses API bridge only showed
examples with models that have `mode: responses` (like o3-deep-research),
which work automatically. This update clarifies that models with
`mode: chat` (like gpt-4o, gpt-5) require the `openai/responses/` prefix
to use built-in tools like web_search_preview.

Changes:
- Explain the `mode` property from model_prices_and_context_window.json
- List models with mode: responses vs mode: chat
- Add example showing the common error and how to fix it
- Add SDK example using the prefix with gpt-4o
- Update proxy example with both automatic and prefix-based configs
- Fix invalid trailing comma in original JSON example
 2026-03-11 13:47:57 -03:00
michelligabriele
24ad510617
feat(mcp): add AWS SigV4 auth support in UI and fix credential merge on edit (#23282) 2026-03-11 09:43:28 -07:00
Chesars
eb30aa34cf chore: restore gpt-4-0613 (still functional despite deprecation date) 
Verified via curl that gpt-4-0613 still responds successfully.
gpt-4-0314 confirmed dead (model_not_found), stays removed.
 2026-03-11 12:31:41 -03:00
Chesars
16de1e300c chore: restore gemini-live-2.5-flash-preview-native-audio-09-2025 (shutdown Mar 19 2026, not yet) 
The shutdown date is March 19, 2026 — we're not there yet.
Restoring both vertex_ai and gemini/ variants.
 2026-03-11 12:24:07 -03:00
Chesars
d43683e875 chore: restore text-embedding-ada-002-v2 (not deprecated) 
text-embedding-ada-002 is not listed as deprecated in OpenAI's
official deprecations page. Re-adding the v2 alias entry.
 2026-03-11 12:22:38 -03:00
Joe Reyna
cbbd51a5ce
fix(codeql): switch to security-extended to fix OOM failures (#23226) 
* fix(codeql): switch to security-extended query suite

The security-and-quality suite produces result sets > 2 GiB on this
codebase, causing fatal OOM failures and blocking CI. Switching to
security-extended reduces query scope to security-only checks, which
still complete successfully. Quality/maintainability checks are
already covered by the existing lint pipeline.

* fix(codeql): exclude OOM queries from security-extended
 2026-03-11 07:38:01 -07:00
Joe Reyna
7d2cc4a3bf
fix(ui): import MCPEvent type into local scope in chat/types.ts (#23330) 2026-03-11 07:37:24 -07:00
Harshit Jain
7db34e3179
Merge pull request #23257 from Harshit28j/litellm_fix-client-close-evict 
fix: fail proxy startup if prisma migrate fails
 2026-03-11 19:45:13 +05:30
Sameer Kankute
d99ddc67d4
Merge pull request #23338 from BerriAI/litellm_fix_multipart_passthrough 
fix(proxy): preserve multipart/form-data boundary in passthrough endpoints
 2026-03-11 19:30:54 +05:30
Sameer Kankute
8e32919074
Merge branch 'main' into litellm_fix_multipart_passthrough 2026-03-11 19:30:45 +05:30
Sameer Kankute
20980f6c26
Merge pull request #23322 from BerriAI/litellm_gemini_embedding_2_support 
[Feat]: Add support for gemini embedding 2 preview
 2026-03-11 19:30:09 +05:30
Chesars
fd46d74424 chore: remove 137 confirmed deprecated/shutdown models from pricing JSON 
Remove models verified as deprecated/shutdown against provider APIs:

- Anthropic (8): Claude 3.5 Sonnet/Haiku (shutdown Feb 2026), Claude 3.7 Sonnet,
  Claude 3 Opus (deprecated)
- OpenAI/Azure (23): GPT-3.5-turbo snapshots, GPT-4 dated snapshots, GPT-4-32k,
  GPT-4.5-preview (shutdown Jul 2025), o1-mini/preview (deprecated Apr 2025),
  old audio/realtime previews, azure/gpt-35-turbo-0301 and -0613 (retired Feb 2025)
- Google PaLM legacy (32): All chat-bison, code-bison, codechat-bison, text-bison,
  textembedding-gecko variants (retired)
- Gemini 1.0/1.5 (28): All variants including gemini/ prefix (shutdown Sep 2025)
- Gemini 2.0 experimental (13): flash-exp, thinking-exp, pro-exp, live-preview,
  image-generation preview (all expired)
- Gemini 2.5 dated previews (14): preview-03-25 through 06-05, flash-image-preview
  (shutdown Jan 2026), live-audio preview (shutdown Mar 2026)
- Veo 3.0 previews (4): shutdown Nov 2025, replaced by GA/3.1
- Perplexity legacy (5): llama-3.1-sonar-* (replaced by sonar-pro/sonar)

Verified sources:
- https://developers.openai.com/api/docs/deprecations/
- https://platform.claude.com/docs/en/about-claude/model-deprecations
- https://ai.google.dev/gemini-api/docs/deprecations
- https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirements

Models intentionally kept:
- All GPT-5 family, azure/gpt-35-turbo-1106 (not confirmed retired)
- gemini-2.5-flash/pro GA, active previews (tts, computer-use, flash-lite, flash-image)
- imagen-3.0-generate-002, all Gemini 3.x, all Claude 4.x
- cerebras/zai-glm-4.6 (active)
- All Mistral, Nebius, Dashscope, Fireworks, Together AI models
 2026-03-11 10:59:35 -03:00
michelligabriele
db4cd87979
docs(web_fetch): add newer Claude models to supported models list (#23251) 
Add Claude Opus 4.6, Sonnet 4.6, Opus 4.5, Sonnet 4.5, and Haiku 4.5
to the web fetch supported models documentation. These models were
missing from the list despite supporting the web_fetch tool.
 2026-03-11 19:09:28 +05:30
Sameer Kankute
ff2fe96717
Merge pull request #23276 from BerriAI/litellm_oss_staging_03_10_2026 
Litellm oss staging 03 10 2026
 2026-03-11 18:54:38 +05:30
Sameer Kankute
f243e5615f
Merge branch 'main' into litellm_oss_staging_03_10_2026 2026-03-11 18:50:03 +05:30
Sameer Kankute
2343149f2d
Merge pull request #23163 from BerriAI/litellm_oss_staging_03_04_2026 
Litellm oss staging 03 04 2026
 2026-03-11 18:46:51 +05:30
Sameer Kankute
43217c8a4b
Merge branch 'main' into litellm_oss_staging_03_10_2026 2026-03-11 18:32:17 +05:30
Sameer Kankute
3dab62023c Merge branch 'main' into litellm_oss_staging_03_04_2026 2026-03-11 18:31:20 +05:30
Sameer Kankute
c2fca1124b fix(proxy): preserve multipart/form-data boundary in passthrough endpoints 
Fixes issue where multipart file uploads through passthrough endpoints failed with RequestValidationError. The proxy was consuming the request body stream and FastAPI was trying to parse multipart bodies as JSON dicts.

Changes:
- Try JSON parsing first for multipart content-type (handles misconfigured clients)
- Skip multipart parsing if JSON succeeds to avoid stream consumption
- Remove custom_body parameter from endpoint_func to prevent FastAPI auto-parsing
- Check for parsed body before using multipart handler
- Add regression test for multipart boundary preservation

Handles both actual multipart uploads and JSON bodies with incorrect multipart content-type headers.

Made-with: Cursor
 2026-03-11 16:52:02 +05:30
yuneng-jiang
628510d1b5
Merge pull request #23298 from BerriAI/litellm_/silly-pasteur 
[Test] UI - Logs: Add unit tests for view_logs components
 2026-03-11 00:57:26 -07:00
yuneng-jiang
09e2676e85
Merge pull request #23328 from BerriAI/litellm_mcp_recheck_health 
[Feature] UI - MCP Servers: Per-server health recheck
 2026-03-11 00:56:44 -07:00
yuneng-jiang
fdf925a3a3 [Feature] UI - MCP Servers: Add per-server health recheck 
Allow users to recheck health for individual MCP servers by clicking
the health status badge. On hover the badge text changes to "Recheck"
with a refresh icon, and the check runs only for that server.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-11 00:54:25 -07:00
Sameer Kankute
a6fb16aea0
Merge pull request #23103 from netbrah/fix/anthropic-tool-schema-type-enforcement 
fix(anthropic): enforce type:"object" on tool input schemas in _map_tool_helper
 2026-03-11 13:06:59 +05:30
Sameer Kankute
b9a311743f
Merge pull request #23104 from netbrah/fix/anthropic-deduplicate-tool-results 
fix(anthropic): deduplicate tool_result messages by tool_call_id
 2026-03-11 13:04:59 +05:30
yuneng-jiang
4f36d29d0c
Merge pull request #23326 from BerriAI/litellm_mcp_permissions_yj 
[Fix] MCP Key Scope Type Fix
 2026-03-11 00:27:18 -07:00
yuneng-jiang
ff2f96d09e Fix mcp_tool_permissions JSON string deserialization in _resolve_team_allowed_mcp_servers 
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-11 00:23:40 -07:00
yuneng-jiang
b48682f61a
Merge pull request #23323 from BerriAI/litellm_mcp_permissions_yj 
[Feature] MCP Server Team-Scoped Filtering for Key Creation
 2026-03-11 00:20:19 -07:00
yuneng-jiang
860cb17571 Fix cross-team MCP server info disclosure and restricted key bypass 
The GET /v1/mcp/server endpoint allowed any authenticated user to pass
an arbitrary team_id and enumerate another team's MCP server config.
Restricted virtual keys could also use the team_id param to bypass
their access limitations. Add team membership check for non-admins
and block restricted keys from using the team_id filter.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-10 23:48:30 -07:00
Sameer Kankute
e394914d34 Fix code qa 2026-03-11 11:49:02 +05:30