langchain-ai/langchainjs @langchain/openai@1.6.0

on GitHub

Minor Changes

• #11232 663b9a2 Thanks @ccurme! - feat(openai): support explicit prompt cache options and content-block breakpoints

Patch Changes

• #11735 5d509ad Thanks @thushanth-bengre-langchain! - fix(openai): send promptCacheRetention: "in-memory" as "in_memory" (the API rejects the hyphenated value), and apply promptCacheKey/promptCacheRetention in the same order as promptCacheOptions: per-call option, then modelKwargs, then the constructor field. On Chat Completions, modelKwargs values are no longer dropped and now take precedence over the constructor field.