Comment by OpenAI

The Safety Hub provides public access to safety evaluation results for our models. [...] evaluations of models under our Preparedness Framework prior to their deployment. Unverified source (2025)