[BUG] Misleading error message when context window is exceeded for third-party models (mimo-v2.5-pro)

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

When using a third-party model mimo-v2.5-pro[1m] via custom API endpoint, once the conversation context grows too large (approaching the model's 1M token limit), Claude Code shows a misleading error:

There's an issue with the selected model (mimo-v2.5-pro[1m]). It may not exist or you may not have access to it. Run /model to pick a different model.

This is clearly a context length overflow, not a model access issue. The error message is confusing and sends users down the wrong debugging path.

What Should Happen?

Claude Code should recognize context overflow errors (HTTP 400/413 or equivalent error codes from the API) and either:

• Show an accurate message like "Context window exceeded for this model"
• Trigger automatic context compression/summarization
• Suggest starting a new conversation

Error Messages/Logs

The error is misclassified as a model-not-found / permission issue, which is completely unrelated to the root cause.

Steps to Reproduce

1. Set model to mimo-v2.5-pro[1m] via /model
2. Have a long conversation until context approaches 1M tokens
3. Observe the misleading error message

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.139

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_