mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-22 00:31:44 +00:00
Adds caching support for the anthropic_messages call type (/v1/messages endpoint), enabling response caching for both streaming and non-streaming requests via the existing litellm.cache infrastructure. Changes: - Add 'anthropic_messages' to CachingSupportedCallTypes - Include Anthropic-specific params (system, stop_sequences, top_k) in cache key - Add cache read/replay branch in _convert_cached_result_to_model_response - Add streaming cache write in AgenticAnthropicStreamingIterator._persist_to_cache() - Comprehensive test coverage (24 tests) Fixes #25653 |
||
|---|---|---|
| .. | ||
| actions/helm-oci-chart-releaser | ||
| codeql | ||
| ISSUE_TEMPLATE | ||
| observatory | ||
| screenshots | ||
| scripts | ||
| workflows | ||
| dependabot.yaml | ||
| deploy-to-aws.png | ||
| FUNDING.yml | ||
| pull_request_template.md | ||
| template.yaml | ||