List what is actually available: GET /v1/models against kRouter returns every model your connected providers serve, with the exact routing id. Check the alias prefix — models are addressed as provider/model, and the wrong prefix routes to a provider that has never heard of the name. If the model was retired, kRouter tells you so by name and points at the live replacement rather than passing the provider's opaque error through. Reconnect the provider if the model should exist on your plan but does not appear.

Reconnecting the provider. — New credentials do not create a model that was retired or renamed. Guessing a close variant of the name. — Ask instead: GET /v1/models returns the exact ids your connected providers actually serve.

All providers

"Model not found" / "Requested entity was not found" — fixing a bad model id

Matches these upstream messages

model_not_found
requested entity was not found
not_found
404 page not found
the model `

What you actually see

{"error":{"message":"The model `gpt-5-turbo` does not exist or you do not have access to it.","type":"invalid_request_error","code":"model_not_found"}}

What does "model_not_found" mean?

The provider does not recognise the model id you sent. Nothing is wrong with your account or your quota — the name is wrong, retired, or not available on your plan.

Why does it happen?

Model ids change more often than people expect: providers rename them, retire preview builds, and gate some behind specific plans. A hard-coded id in an old config is the usual culprit.

How do I fix it?

List what is actually available: GET /v1/models against kRouter returns every model your connected providers serve, with the exact routing id.
Check the alias prefix — models are addressed as provider/model, and the wrong prefix routes to a provider that has never heard of the name.
If the model was retired, kRouter tells you so by name and points at the live replacement rather than passing the provider's opaque error through.
Reconnect the provider if the model should exist on your plan but does not appear.

Is it really this error?

These get mistaken for each other, and chasing the wrong one is where the time goes.

Looks like: An authentication problem, since 'do not have access' sounds like permissions.

Actually: Providers use one message for both. If other models work on the same key, it is the id — not your access.

Looks like: A kRouter routing bug.

Actually: Usually the alias prefix: provider/model means the prefix decides who receives the name. The wrong prefix sends a valid id to a provider that never had it.

What will not fix it

Reasonable-sounding fixes that do not apply here — worth ruling out before you spend an afternoon on them.

Reconnecting the provider.
New credentials do not create a model that was retired or renamed.
Guessing a close variant of the name.
Ask instead: GET /v1/models returns the exact ids your connected providers actually serve.

What kRouter does automatically

kRouter parks that model on that account for 30 minutes. It does NOT try your other accounts, because the same request would fail identically on all of them — the error is returned to you straight away instead of burning every credential you have.

Stop hitting this error

kRouter routes your AI coding tools across every provider and account you own, so a limit on one does not stop your work. It is free, MIT-licensed and runs on your machine.

Install kRouter