Context caching with Gemini is so good! Here I am caching the entire Gemini Cookbook (around 400k tokens) as an insanely long prompt to create the best Gemini app developer on the planet. Watch Gemini answer any coding questions related to its own APIs.

And here is all the code needed!

🚀 Don't gamble with your portfolio! Use our advanced hybrid quant risk tool using on/off-chain data and make informed decisions. 📈 Acess to 1000+ charts for your crypto journey. 📚Join our Premium Telegram for daily alerts. 📊+21 projects supported. 🏗️ Beginners and experts.

This opens an entire new world for AI applications. Great work to @OfficialLoganK, @joshtwoodward, @JeffDean and the rest of Gemini team!

This is so cool. I can’t wait to see what else people cook up with long context and flash : )

Does it charge for the cached tokens? And this just speeds it up? Or are there cost savings too?

It charges for cached tokens and there is no limit of how long you store the tokens.

Can't wait to get access to the 2M. Still on the waitlist.

Nice! Did you convert the entire cookbook into JSON? 😅

I used my own repo :)

Awesome this is a striking demo! However it seems they are charging by the hour for context caching. Nothing exorbitant for a few hours but it can't remain in the cache forever

Still have a healthy skepticism about needle finding in a haystack that big, but this potentially could be a huge leap forward in consumer models. Alleviates some of the pressure that's been necessitating workarounds for context window injection / medium term memory like RAG