Hi everyone,
I’m experiencing an issue in Firebase Studio when using Gemini from the chat/prototyper interface.
Whenever I send a prompt, I receive the following error:
Sorry, I hit a snag. Please try again shortly or modify your prompt.
[GoogleGenerativeAI Error]:
Error fetching from
https://monospace-pa.googleapis.com/v1/models/gemini-3-flash-preview:streamGenerateContent?alt=sse:
[400 Bad Request]
The input token count exceeds the maximum number of tokens allowed 1048576.
What confuses me is that I already significantly reduced my context backup file:
context_backup/capra-context-state-20260526184319.json
Originally it contained over 10,000 lines, but I reduced it to roughly 400 lines. Despite this, Firebase Studio continues to return the exact same token limit error.
A few questions:
-
Does Firebase Studio cache previous conversation context somewhere else besides the context backup file?
-
Is there another memory, chat history, workspace index, or hidden context that Gemini is still loading?
-
How can I determine what content is actually being sent to the Gemini model?
-
Is there a way to completely reset the AI context for a project/workspace?
Any guidance would be appreciated.
Thanks!