Skip to main content
LiteLLM is moving to Rust Read the latest updates.

LiteLLM v1.109.3 and v1.106.12 fix GPT-6.2 caching on stored requests

Released

Deploy this version​

docker run \
-e LITELLM_MASTER_KEY=sk-<paste-a-long-random-key> \
-e DATABASE_URL=postgresql://<user>:<password>@<host>:5432/<dbname> \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
docker.litellm.ai/berriai/litellm:1.109.3

This release is published as ghcr.io/berriai/litellm:v1.109.3. See the GitHub release and the full releases page.

v1.109.3 is a patch release on top of v1.109.2. There are no new database migrations. The v1.109.3 tag points at 37c2769.

What's Changed​

  • Caching: Fixed GPT-6.2 system prompts missing the prompt cache on /v1/chat/completions and /v1/messages unless store: false - PR #52906
  • Keys, teams and auth: Fixed JWT key mappings without jwt_issuer taking 5 minutes to apply when litellm_jwtauth.issuers is set, broken since v1.103.0 - PR #52350
  • Keys, teams and auth: Fixed /key/info letting a key with no user read other userless keys by hash or alias - PR #52838 and PR #52925
  • Deployment: Changed filelock to 4.2.4 and soupsieve to 4.1 to match main - PR #52937
  • Deployment: Changed the Docker images to a newer Wolfi base image - PR #53008

The changes are backported in PR #52712, PR #52929 and PR #52937.

Full Changelog​

https://github.com/BerriAI/litellm/compare/v1.109.2...v1.109.3

LiteLLM Enterprise
SSO/SAML, audit logs, spend tracking, multi-team management, and guardrails, built for production.
Learn more →