36 ms·

How many tokens are you burning daily?

• gls2ro 8mo ago
  Not the OP but I think in case of scanning and tagging/summarization you can run a local LLM and it will work with a good enough accuracy for this case.
• storystarling 8mo ago
  The real cost driver with agents seems to be the repetitive context transmission since you re-send the history every step. I found I had to implement tiered model routing or prompt caching just to make the unit economics work.