v1.86.6 - DB Resilience, Passthrough & Dependency Backports
Deploy this version​
- Docker
- Pip
docker run \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
docker.litellm.ai/berriai/litellm:1.86.6
pip install litellm==1.86.6
v1.86.6 is a patch release on top of v1.86.5. It brings the 1.84.8 database-resilience and passthrough set onto the 1.86 line, adds the budget-reservation toggle, hardens Anthropic streaming logging, and refreshes dependencies.
What's Changed​
- fix(router): use forwarded model_id for native Azure container IDs - PR #27921
- fix(proxy): expose Prisma idle/connect timeout and extra DB URL params - PR #28395
- feat(proxy): add
disable_budget_reservationgeneral setting - PR #29493 - fix(proxy): recover from cached-plan errors by reconnecting the Prisma client - PR #29983
- feat(proxy): add option to disable server-side prepared statements for DB lookups - PR #29984
- fix(proxy): return 5xx on DB infra errors during auth; reserve 401 for genuine auth failures - PR #29986
- fix(passthrough): resolve costing model when body model is unknown - PR #30160
- fix(passthrough): skip
[DONE]sentinels and non-JSON SSE frames in Anthropic streaming logging - PR #30202 - fix(proxy): return deprecated-key lookup result directly in get_data combined view - PR #30327
- chore(deps): bump pypdf, tornado, the aiohttp constraint, vitest, and brace-expansion - PR #30220
Full Changelog​
https://github.com/BerriAI/litellm/compare/v1.86.5...v1.86.6