[FEATURE]: Improve Auto Router Error Messaging and Model Status Indicators

When using Auto Routers, the current error message from the provider console is not sufficiently informative:

Error from provider (Console): Upstream request failed: [rate_limit_exceeded] Ra...

This message does not help users understand which model the Auto Router attempted to connect to, nor does it provide retry context.

Proposed Improvements

Example format:
Trying ModelName (free) ... Attempt #1 Retry in 30s

This allows users to see which model is being attempted and the retry schedule.

Expired Token Highlighting

In the model list UI, change the font color to red for models where token usage has expired.
This provides immediate visual feedback so users can manually select valid models without trial and error.

Benefits

@davidjeba

Thanks for the suggestion, @davidjeba! As I understand it, you'd like two things:

When an Auto Router request fails, the error says which model was tried and gives retry details (attempt number, time until the next retry), instead of the raw Upstream request failed: [rate_limit_exceeded] … text.
In the model list, models whose free usage has run out are marked (e.g. in red), so you can pick one that works without trial and error.

Maintainers will look at this as a feature request.

Related issues you may want to follow:

[FEATURE]: Configurable model fallback on transient errors and timeouts #48991: configurable fallback to another model on transient errors and timeouts
feat(core): make the model retry budget configurable via env #59448: making the model retry budget configurable
feat(core): read OpenCode Console usage limits and classify budget refusals #58694: reading OpenCode Console usage limits and classifying budget refusals, which an availability indicator could build on

Navigation Menu