BerriAI/litellm v1.39.6

on GitHub

We're launching team member invites (No SSO Required) on v1.39.6 🔥 Invite team member to view LLM Usage, Spend per service https://docs.litellm.ai/docs/proxy/ui

👍 [Fix] Cache Vertex AI clients - Major Perf improvement for VertexAI models

✨ Feat - Send new users invite emails on creation (using 'send_invite_email' on /user/new)

💻 UI - allow users to sign in with with email/password

🔓 [UI] Admin UI Invite Links for non SSO

✨ PR - [FEAT] Perf improvements - litellm.completion / litellm.acompletion - Cache OpenAI client

What's Changed

• Fix warnings from pydantic by @lj-wego in #3670
• Update pydantic version in CI requirements.txt by @lj-wego in #3938
• Allow admin to give invite links to others by @krrishdholakia in #3875
• Update model config definition to use v2 style by @lj-wego in #3943
• Add OIDC + unit test for bedrock httpx by @Manouchehri in #3688
• (fix) Update Mistral model list and prices by @alexpeattie in #3945
• feat - send_invite_email on /user/new by @ishaan-jaff in #3942
• [UI] Admin UI Invite Links for non SSO users by @ishaan-jaff in #3950
• [Feat] Admin UI - invite users to view spend by @ishaan-jaff in #3952
• UI - allow users to sign in with with email/password by @ishaan-jaff in #3953
• feat(proxy_server.py): add assistants api endpoints to proxy server by @krrishdholakia in #3936
• [Fix] Cache Vertex AI clients - Perf improvement by @ishaan-jaff in #3935
• fix(bedrock): convert botocore credentials when role is assumed by @pharindoko in #3939

New Contributors

• @lj-wego made their first contribution in #3670
• @alexpeattie made their first contribution in #3945
• @pharindoko made their first contribution in #3939

Full Changelog: v1.39.5...v1.39.6

Docker Run LiteLLM Proxy

docker run \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
ghcr.io/berriai/litellm:main-v1.39.6

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