OpenAI introduces GPT-6.2 Astra and Sol
Original title: Introducing GPT-6.2 Astra and Sol
Astra is priced 20% below GPT-6 Astra in the API and leads several professional benchmarks at a fraction of rival cost.
OpenAI released GPT-6.2 Astra for complex professional work and GPT-6.2 Sol for everyday tasks. Both are described as using fewer output tokens than prior models. In the API, Astra input and output prices are 20% lower than GPT-6 Astra, at $8, $0.40 cached input, and $40 per million tokens; Sol keeps GPT-6.1 Sol prices. OpenAI says Astra leads Agents’ Last Exam, Terminal-Bench-Science 0.1, and the Vals Science Index.
Unconfirmed: Benchmark leads and cost comparisons are OpenAI’s claims; Fable 5.2 leaderboard results exclude nine fallback tasks.
- Astra scores 5.2 points above GPT-6 Astra on Agents’ Last Exam at under one-third the cost.
- Astra is 5.5 points ahead of Claude Fable 5.2 at 5% of the cost, OpenAI says.
- Terminal-Bench-Science 0.1 score is 9 points above GPT-6 Astra.
- Sol stays within 0.5 points of Astra on FrontierCode 2.0 at about one-fifth the cost.
From the original article
13 min read at openai.comIntroducing GPT‑6. 2 Astra and Sol, with better professional outputs, coding, computer use, and efficiency at lower cost across ChatGPT Work, Codex, and the API.
openai.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.