Constant "Too many requests" and "Invalid model" errors regardless of model selection

[soft salmon]

@carmine osprey

[bright bear]

hello i am using the kiro ai agent but it's been a while that is not working and give me this message " An unexpected error occurred while processing your input." what i have to do

[candid notch]

Hey there! These are different class of errors:

1. The "Too many requests" error can occur if a large amount of requests are made from the Kiro client in a short duration span.
2. The "Invalid model" error can occur due to model availability in certain geographic locations as outlined by model providers - https://kiro.dev/docs/models/#model-availability.
3. If you run into the "unexpected error message" verify the following:
   • Incorrect mcp tool schema being passed, temporarily disable mcp servers to verify
   • Adding large documents like xls to the chat input, use the restore button or checkpointing feature to rollback to a previous input, remove the file and retry

If you rule these out, capture the browser developer logs and share them in this thread alongside the prompt.

[steady island]

Oook))) Invalid model. That may well be the case, but how are geographical features related to the fact that the one model is sometimes available and sometimes not?

[candid notch]

If the model provider doesn't support a specific geographic location those models will not appear in the clients. Are you observing otherwise? What model are you referring to?

[steady island]

For example, Opus 7. Sometimes the code is generated, but other times I encounter an error nvalid model. However, I am not changing the region.

[candid notch]

Just to confirm Opus 4.7 and are you logging in with IDC?

[steady island]

No, actually—I log in using the standard GitHub authentication.

[candid notch]

Opus 4.7 is not yet available to individual users. We are working on making it available to all Kiro subscription tiers soon.