OpenAI Releases GPT-5.2 and ChatGPT Health
TL;DR
OpenAI has unveiled GPT-5.2 with major improvements in general intelligence, long-context understanding, agentic tool-calling, and vision. GPT-5.2 Thinking performs at/above human expert level on 70.9% of knowledge work comparisons. Simultaneously, ChatGPT Health was launched – a dedicated space for health conversations with secure connections to Apple Health and wellness apps. Health data is isolated and not used for training. Also new: GPT-5.2-Codex for coding tasks.
Key Points
- OpenAI has unveiled GPT-5
- 2 with major improvements in general intelligence, long-context understanding, agentic tool-calling, and vision
- 2 Thinking performs at/above human expert level on 70
- 9% of knowledge work comparisons
- Simultaneously, ChatGPT Health was launched – a dedicated space for health conversations with secure connections to Apple Health and wellness apps
Summary
OpenAI has unveiled GPT-5.2 with major improvements in general intelligence, long-context understanding, agentic tool-calling, and vision. GPT-5.2 Thinking performs at/above human expert level on 70.9% of knowledge work comparisons. Simultaneously, ChatGPT Health was launched – a dedicated space for health conversations with secure connections to Apple Health and wellness apps. Health data is isolated and not used for training. Also new: GPT-5.2-Codex for coding tasks.
Nauti's Take
GPT-5.2 isn't an incremental update – it's a statement. The numbers are impressive: 70.9% human expert level on knowledge work. But the real bombshell is ChatGPT Health. 230 million users ask about health weekly? OpenAI is positioning itself as a serious healthcare player. The strict data separation (no training data from health area) was smart – without it, the FDA discussion would be inevitable. GPT-5.2-Codex will land in our daily work. Our take: this is OpenAI's answer to Claude – and it's strong.
Frequently Asked
What is OpenAI Releases GPT-5.2 and ChatGPT Health?
OpenAI has unveiled GPT-5.2 with major improvements in general intelligence, long-context understanding, agentic tool-calling, and vision. GPT-5.2 Thinking performs at/above human expert level on 70.9% of knowledge work comparisons. Simultaneously, ChatGPT Health was launched – a dedicated space for health conversations with secure connections to Apple Health and wellness apps. Health data is isolated and not used for training. Also new: GPT-5.2-Codex for coding tasks.
Why does this matter?
OpenAI has unveiled GPT-5
What are the key takeaways?
OpenAI has unveiled GPT-5. 2 with major improvements in general intelligence, long-context understanding, agentic tool-calling, and vision. 2 Thinking performs at/above human expert level on 70