OpenAI - Helicone

ok cool -- I'll consider using it, for now I decided to just set the baseurl on the default openai package, and added the headers myself

would caching work if I added the headers myself? or is there something fundementally missing that needs to be added on your side to enable caching?

we perform caching within our cloudflare worker

so its only on the proxy with the cache header

got it, and does the js library use the proxy?

no it does not use the proxy, it writes directly to our servers

and log asynchronously

gotcha, sounds good

thanks for the info!

Hi @abe sorry for the confusion. Our package does support proxy if you want to use it.

https://www.npmjs.com/package/helicone

(The package you shared is depricated, we need to remove it)

It does not support Fetch yet and the OpenAI V4 package. We are adding it this next sprint 🙂

Please let me know if you have any questions.

@abe I updated the docs here
https://docs.helicone.ai/features/advanced-usage/caching