Optimizing Token Usage for Large Inputs in Bolt.diy

Hi everyone,
I’m a fan of Bolt.diy and have been exploring it with my own custom model. While the model supports a maximum of 35,000 tokens, I often encounter token overflow issues when uploading folders and querying the model.

Does anyone have suggestions or strategies for:
Reducing token usage when working with large inputs?

1 Like

did you get any response on this?

** @leex279 any input or thoughts on this?**

hey, this is a pretty old topic from a time, where Context Optimization (Settings → Features → Context Optimization) was not present.
So normally there shouldnt be much problems with this stuff, when using proper Models.

If you still got problems, please open a new topic with a chat export and more information/details (Provider, Model, etc.).

Please see
enjoy this emoji
:man_cook:
“Proposal for
Settings → Features → Append WebContainer (Package) To (The End Of) Prompts”