What
When you pick a model with /model, Claude Code checks the name with the API (Anthropic's service). If that check hits a generic API error, the message now uses a cleaned-up error text, drops trailing punctuation and ends with · model not changed.
The result is also marked retryable, meaning trying again may work, when:
- there is no status code
- the status is 408 (timeout)
- the status is 429 (too many requests)
- the status is any 5xx (server-side error)
Failures that are not API errors, shown as 'Unable to validate model', are now always marked retryable.
Why
The message now states plainly that your model setting stayed as it was, and temporary problems are marked as worth retrying.
Names in the bundle/model
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Added since
A small documentation edit on Claude Code on Amazon Bedrock touched a line naming /model after this was published.
The [setup wizard](#sign-in-with-bedrock) offers a 1M context option when it pins models. To enable it for a manually pinned model instead, append `[1m]` to the model ID. See [Pin models for third-party deployments](/docs/en/model-config#p…amazon-bedrock see the edit
Confirmed since
Anthropic's documentation has since written up /model, on Configure cloud environments.
| Your organization's [server-managed settings](/docs/en/server-managed-settings) | Yes, except in [Claude Tag](https://claude.com/docs/claude-tag/overview) sessions | Fetched from Anthropic's servers when the session starts. See [Surface …cloud-environments see the edit
Two sources agreeTwo things we can check say the same as this entry.
Anthropic's documentation agrees
Anthropic's documentation has since written up /model, on Configure cloud environments.
Anthropic's release notes agree
Fixed /model showing the raw API error JSON and request ID when the API refuses the picked model; it now shows the server's message and says…