BerriAI/litellm v1.74.2-nightly

on GitHub

What's Changed

• add grok-4 configs to table by @fcakyon in #12476
• fix slack alerts by @jugaldb in #12464
• Add devstral-small-2507 and devstral-medium-2507 models by @xingyaoww in #12484
• [Bug Fix] fix parsing environment_variables from config.yaml (arize logger integration fix) by @ishaan-jaff in #12482
• [Chore] Don't emit warning for Max in memory queue flush count by @ishaan-jaff in #12489
• Add Azure OpenAI o3-deep-research model pricing support by @neubig in #12493
• Feat(bedrock): support api key authentication for AWS Bedrock API by @ishaan-jaff in #12495
• Added validate payload error by @jugaldb in #12494
• [MCP Gateway] - Add custom cost configuration for each MCP tool by @ishaan-jaff in #12499
• [Feat] Add support for editing MCP cost per tool by @ishaan-jaff in #12501
• [docs]: Fix typo and import required types for proxy call hooks by @Rayshard in #12487
• fix: handle reasoning parameters and response in responses bridge by @aholmberg in #12433
• Added dashscope (alibaba's cloud - qwen) as a provider by @minghao51 in #12361
• feat: improve user dropdown UI with premium badge and cleaner layout by @colesmcintosh in #12502

New Contributors

• @Rayshard made their first contribution in #12487
• @minghao51 made their first contribution in #12361

Full Changelog: v1.74.1-nightly...v1.74.2-nightly

Docker Run LiteLLM Proxy

docker run \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
ghcr.io/berriai/litellm:main-v1.74.2-nightly

Don't want to maintain your internal proxy? get in touch 🎉

Hosted Proxy Alpha: https://calendly.com/d/4mp-gd3-k5k/litellm-1-1-onboarding-chat

Load Test LiteLLM Proxy Results