Skip to main content

v1.89.5 - Vertex Batch Upload Fix & Key Redaction

Update: no performance regression found

An earlier version of this note flagged a potential throughput regression. We investigated and could not confirm or reproduce any regression in the released version. The one report we received came from a deployment running custom code on top of what we shipped, and our testing points to those changes, not LiteLLM, as the likely cause.

Correctness and error rates were never affected. If you're on this version, there's nothing you need to do.

We're still monitoring incoming reports and will update this note if anything changes.

Deploy this version​

docker run \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
docker.litellm.ai/berriai/litellm:1.89.5

v1.89.5 is a patch release on top of v1.89.4. It switches Vertex AI batch-file uploads to a single media upload so large uploads no longer fail with 499 errors, and redacts the API key from /key/info client error messages.

What's Changed​

  • fix(vertex_ai/files): single media upload for batch files to fix 499s on large uploads - PR #31653
  • fix(proxy/client): redact api key from key/info client error messages - PR #31342

Full Changelog​

https://github.com/BerriAI/litellm/compare/v1.89.4...v1.89.5

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