Comment on Prompt caching for cheaper LLM tokensparentComments−moebrowne8mohttps://platform.openai.com/docs/guides/prompt-caching#requi...Caching is available for prompts containing 1024 tokens or more.No mention of caching being in blocks of 1024 tokens thereafter.−IanCal8moAt launch it was described as being in blocks of 128https://openai.com/index/api-prompt-caching/
Comments
https://platform.openai.com/docs/guides/prompt-caching#requi...
No mention of caching being in blocks of 1024 tokens thereafter.
At launch it was described as being in blocks of 128
https://openai.com/index/api-prompt-caching/