Caching and cache misses for requests with new seeds

cache is on headers + body, if you change anything in that it'll miss
there's semantic cache as well is that what you're looking for? https://portkey.ai/docs/product/ai-gateway/cache-simple-and-semantic#semantic-cache

Semantic cache does not work for use since we are processing tabels, and change in a single value needs to retrigger calculation.

then what type of cache are you talking about if your body is not the same with every request 🥲

If you're thinking about provider specific caching like openai caching, that is provider specific, for example openai caches any context which is > 1024 tokens

for anthropic you have to send cache key in the specific message content item

I just removed the seed parameter to fix it