Step 1
Confirm the exact model identifier
Search the request path, environment variables, wrappers, defaults, test fixtures, and provider configuration. Many applications keep the model name somewhere other than the call site.
OpenAI API error
OpenAI model_not_found: what to check before you swap the model ID
A model_not_found failure can mean the requested model is unavailable to the project, the identifier is wrong, or a deprecated model has reached shutdown. If the model was retired, changing the string is only the first migration step.
What to check
Step 2
Check the official shutdown schedule
OpenAI says a model is no longer accessible after its shutdown date. Compare the exact identifier against the current deprecation table before choosing a replacement.
Step 3
Review request compatibility
A newer model may use different reasoning, tool-calling, sampling, or token parameters. Re-run representative requests instead of assuming an identifier-only change is safe.
If this appeared after a previously working deployment, treat the model identifier and shutdown schedule as the first two checks, then verify that your replacement request shape is supported.
Check the rest of the repository
Fixing the first error does not prove the rest of the codebase is clear. Sunset scans source text for scheduled OpenAI shutdown risk and shows severity before you pay.