When listing models for v1beta api version, “createCachedContent” method is only available within 001 models, both flash and pro.
When this functionality will be rolled out for new 002 models?
2 Likes
Hi @Pavel_Mikhaylov.
Welcome to the forum. I already escalated this to the internal team. Soon, will get the clarity on this.
Thanks.
2 Likes
One of my use cases highly depends on it, thanks!
Similar with flash-8b
Hi! Any updates there?
Hi @Pavel_Mikhaylov . Not yet. We didn’t hear back on this. Will follow up again with the team.
Context caching for the 002 models is supported now. 
3 Likes
Hey Govind, I have been playing with 002 + cache and I noticed that the new search grounding does not work (flash/pro 002 + google search + using cache). 001 + google search + cache is fine for example, and so is 002 + google search or 002 + cache.