By simply looking at the response from GET /v1/models, you can now determine whether a model accepts thinking: {type: ...