GPT-5.5 Instant
GPT-5.5 Instant becomes ChatGPT's default model for everyone, with 52.5% fewer hallucinated claims than GPT-5.3 Instant on high-stakes prompts.
Faster everyday model with tighter answers, a more natural tone and better use of shared context and memory. Factuality gains concentrate in medicine, law and finance prompts. The API chat-latest alias tracks the model used in ChatGPT.
- Date
- Tuesday, 5 May 2026
- Lab
- OpenAI
- Kind
- model
- Access
- closed API
Figures
| Measure | Value | Measured by |
|---|---|---|
| Hallucinated claims vs GPT-5.3 Instant (high-stakes prompts) | 52.5% fewer internal evaluation | company |
GPT-5.5-Cyber went to limited preview on 2026-05-07 for defenders of critical infrastructure under Trusted Access for Cyber.
Sources
- openai.com/index/gpt-5-5-instant/
- deploymentsafety.openai.com/gpt-5-5-instant/
- developers.openai.com/api/docs/changelog
This record was checked against its sources on 6 October 2026. How we check