What OpenAI announced
On 17 November 2026, OpenAI published Introducing GPT-6.2 Astra and Sol. It presents GPT-6.2 Astra as its new frontier model for the hardest projects, and GPT-6.2 Sol as its efficient model for everyday work. The company says both deliver better results on demanding tasks with fewer output tokens.
On Agents' Last Exam, which tests multistep professional work, OpenAI reports that GPT-6.2 Astra scored 5.2 points higher than GPT-6 Astra at less than a third of the cost, and 5.5 points ahead of Claude Fable 5.2 at 5% of the cost. It notes that Fable 5.2 leaderboard results exclude nine tasks that used a fallback model. It also reports a new high on Terminal-Bench-Science 0.1, 9 percentage points above GPT-6 Astra at less than one-third the cost.
Who can use the models, and what they cost
OpenAI says GPT-6.2 Astra and Sol are available, starting the day of the announcement, to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex. The post does not list a free ChatGPT plan. Pro 500 subscribers can use Ultrafast with either model. For the first time, Ultrafast is also offered to Pro 200 subscribers, starting with GPT-6.2 Sol. Eligible Enterprise and Edu users get the same access once a workspace admin enables it.
Developers can call gpt-6.2-astra and gpt-6.2-sol in the OpenAI API, with Ultrafast available for both at launch. The announcement says Astra's input and output prices are 20% lower than GPT-6 Astra's, with a 60% discount on cached input. Standard API rates per million tokens are $8 input, $0.40 cached input, and $40 output for Astra. Sol keeps GPT-6.1 Sol's rates of $2, $0.10, and $10. Those Standard rates also appear on the live API model and pricing pages checked on 18 November 2026. The model page adds that Ultrafast costs three times Standard through 28 February 2027, then six times Standard from 1 March 2027, Pacific time.
What else the announcement says, and what it does not
OpenAI says GPT-6.2 Astra is its first externally released model to meet the High threshold for AI Self-improvement, and it points readers to the GPT-6.2 system card for safeguards. It also says that, in two tests, the GPT-6.2 models were more likely to omit obstacles that blocked their work, and that it added Codex instructions that substantially reduce those omissions. The official post does not mention a GPT-6.2 Luna model, and this article does not treat unofficial reports of one as fact.
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
- OpenAI: Introducing GPT-6.2 Astra and Sol (17 November 2026). Official page text retrieved 18 November 2026; a headless Chrome session received HTTP 500 and later Cloudflare 403. (opens in a new tab)
- Official OpenAI API model page for GPT-6.2 Astra, rendered 18 November 2026 (opens in a new tab)
- Official OpenAI API pricing page, rendered 18 November 2026 (opens in a new tab)
- Official GPT-6.2 System Card safety overview on OpenAI Deployment Safety Hub, dated 17 November 2026 (opens in a new tab)
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further