GPT-5.1 Instant and GPT-5.1 Thinking System Card Addendum
Key point
OpenAI has released a system card addendum with updated safety metrics for the GPT-5.1 Instant and Thinking models.
Details
GPT-5.1 Instant offers better conversational ability and improved instruction-following compared to previous models, along with adaptive reasoning that lets it judge for itself how long to spend on reasoning. GPT-5.1 Thinking adjusts its thinking time even more precisely to fit the question. GPT-5.1 Auto automatically routes queries to the optimal model without requiring users to select a model themselves.
This addendum provides updated safety metrics for the new model versions. In particular, as part of the pre-deployment safety review, evaluations were expanded to cover mental health (delusions, psychosis, mania, etc.) and emotional reliance (unhealthy attachment to ChatGPT, etc.).
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.