The requested model is not supported.

[balejs]

Describe the bug

Copilot keeps showing this message for any queries. This happens with all models, albeit /model tells me that it have access to:

❯ 1. Claude Sonnet 4.5 (1x) (default)
2. Claude Sonnet 4 (1x)
3. Claude Haiku 4.5 (0.33x)
4. GPT-5 (1x) (current)
5. Cancel (Esc)

Affected version

0.0.351 Commit: 90770bc

Steps to reproduce the behavior

1. launch copilot
2. add a prompt like "double check that the documentation of this project is accurate"

Expected behavior

The model does not throw an error

Additional context

• OS: MacOS 15.3.2
• verbose logs aren't returning anything particular, can probably collect some useful ones if someone gives me the correct command line

[balejs]

The complete error log line is:

Model call failed: {"message":"The requested model is not supported.","code":"model_not_supported","param":"model","type":"invalid_request_error"} (Request ID: C275:33D2F9:E76233:11E11A4:6920044E)

[dariogliendo]

Having the same issue for a few days now, checked my subscription and it's all right, tried different models, it's just dead, considering changing provider soon if it keeps this way :/

[rianlowrey]

+1, for me it started to occur after the trial expired; despite logging back in again to refresh the credentials. I was billed today too, checked the github billing, and it confirmed automatic enrollment in copilot pro for which I was billed. I get the same errors:

 ● Logged in as user: redacted
 ● Connected to GitHub MCP Server

 ● Signed in successfully as redacted! You can now use Copilot.

 > /status

 ✗ Model call failed: {"message":"The requested model is not supported.","code":"model_not_supported","param":"model","type":"invalid_request_error"} (Request ID: ED75:36E635:96024B:D65798:6922540A)

 ✗ Model call failed: {"message":"The requested model is not supported.","code":"model_not_supported","param":"model","type":"invalid_request_error"} (Request ID: ED75:36E635:96088E:D66127:6922540A)

[rianlowrey]

I managed to fix the issue in my case. For some reason https://github.com/settings/copilot/features reset all the "default" models to no value/unset.

The old client: Version 0.0.354 · Commit 076bd172b was giving the unusual error, as noted above.

I manually updated the client side, Version 0.0.362 · Commit 0c95944 to see if the problem was due to a backwards compatibility issue with having the server be at a newer version. After I updated the client, I received a slightly different message:

 > /status

 ✘ Execution failed: No model available. Check policy enablement under GitHub Settings > Copilot

So I checked the settings and discovered that I need to manually enable each model.

After doing so, you need to restart the cli, and the models you enabled should be available.

[balejs]

I’ve just tried @rianlowrey ‘s approach and can
confirm that it worked for my case, too. All models are now correctly accessible.