105532666284975590557Q: now it's sonnet 3.5 ;) cool
Seriously sonnet 3.5 is nice. 200k contexte ...
Maybe you can try gemini pro with very high contexte 2000k contexte !
I am asking myself how compare with you deep dive emedding process as i am asking myself how you have script compression and embedding and semantique search together. compare with full loading in contexte the whole documents and let's the llm embedding doing the job !
If you offer BYOK let us make infinite loop and deep dive ... apprecate better local embedding too ...
Jon_Socrates
Edited Nov 3, 2024A: BYOK will be shipped, it's highly requested! Note that BYOK won't affect Deep Dive limits because the Deep Dive costs are not AI models, but our proprietary software.
Noted on local embeddings, always looking for a better local model to include!