gpt-5.4-mini works fine with the Responses API but fails when used through the Batch API. The batch endpoint resolves it to gpt-5.4-mini-2026-03-17-batch, which returns a model_not_found error.
{
"error": {
"message": "The requested model 'gpt-5.4-mini-2026-03-17-batch' does not exist.",
"type": "invalid_request_error",
"param": "model",
"code": "model_not_found"
}
}
I’ve got gpt-5.4-mini on the code, the date and batch were added automatically on the OpenAI API