inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
Why it matters
Requests carrying tools now route on live quality signals instead of a hand-maintained provider list, which matters most in the launch window when inferenceRunning a trained model to get answers — the phase where AI is actually used, as opposed to trained.Full definition → engines have not yet stabilized on a new model's chat template and parameters.
