OpenAI releases GPT-6.2 Astra and Sol, ranked #2 on Intelligence Index
Start here
OpenAI released GPT-6.2 Astra and GPT-6.2 Sol, available in ChatGPT Work, Codex and the API as gpt-6.2-astra and gpt-6.2-sol. On Artificial Analysis's Intelligence Index, Astra scores 62 and Sol 57, both behind only Anthropic's Claude Fable 5.2 at 65; Astra gains 5 points over GPT-6 Astra and Sol gains 2. Artificial Analysis places Sol on the Pareto frontier of Intelligence Index versus cost per task. Astra is priced at $8/$40 per million input/output tokens, 20% below GPT-6 Astra, with cached input cut to $0.40 from $1.00; Sol stays at $2/$10. Requests over 272K input tokens double input and cache rates and cost 1.5x for output, and Ultrafast mode is 3x standard through February 28, 2027, then 6x. Coding agent scores did not improve: Astra is flat and Sol loses 3 points, driven by a 14-point drop in SWE-Atlas-QnA-AA. OpenAI says Astra roughly doubles Claude Fable 5.2's pass rate on Harvey LAB-AA at about one-sixth the cost, while Epoch AI found Astra barely improved over GPT-6 Astra on real-world research tasks. Astra also posts a 59% AA-Omniscience hallucination rate, and Simon Willison reports Astra trailing Fable 5.2 on several benchmarks.
Generated from reports · updated 39 min ago
Developments
3 developments
- November 18 03:09 · 1 reportsGPT-6.2 Sol tops Artificial Analysis score-vs-cost frontierHacker News · AI:GPT-6.2 Sol moves the Pareto frontier in Artificial Analysis score vs. cost
- November 18 01:45 · 4 reportsOpenAI releases GPT-6.2 Astra and SolOpenAI News:OpenAI releases GPT-6.2 Astra and Sol
- November 17 00:00 · 1 reportsGPT-6.2 Astra and Sol rank #2 on Intelligence IndexArtificial Analysis Articles:OpenAI releases GPT-6.2 Astra and GPT-6.2 Sol, taking #2 on Artificial Analysis Intelligence Index
Report timeline
Follow the reports to see the event from different sides.
- Simon WillisonPicksOpenAI releases GPT-6.2 Astra and GPT-6.2 Sol, tested by Simon Willison
OpenAI released GPT-6.2 Astra, a frontier model for hard projects, and GPT-6.2 Sol, an efficient model for everyday work, both available in the API as gpt-6.2-astra and gpt-6.2-sol. Simon Willison shipped LLM 0.38 with support for the models and notes Astra costs $8/MTok input and $40/MTok output, 80% of GPT-6 Astra's price, with cached input at $0.40/million versus $1/million before. He reports mixed early reception: Astra trails Claude Fable 5.2 on several benchmarks, and neither new model morphed Yonkers into a mech as well as Fable 5.2 or Gemini 4 Carbon in his Codex CLI animation test.
- MarkTechPostPicksOpenAI Releases GPT-6.2 Astra and GPT-6.2 Sol, With Astra API Prices Down 20%
OpenAI released GPT-6.2 Astra and GPT-6.2 Sol, two closed models available in the API as gpt-6.2-astra and gpt-6.2-sol, with Astra's input and output prices 20% below GPT-6 Astra at $8 and $40 per 1M tokens. Both models share a 1,050,000-token context window with 128,000 max output, and Astra scores 62 on the Artificial Analysis Intelligence Index, second to Claude Fable 5.2 at 65, while Sol scores 57. The report notes coding agent scores did not improve and Sol slipped 3 points, and Astra posts a 59% AA-Omniscience hallucination rate.
- The DecoderPicksOpenAI launches GPT-6.2 Astra and Sol with lower API prices and benchmark gains
OpenAI released GPT-6.2 Astra for demanding professional work and GPT-6.2 Sol for everyday use, both available in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users and through the API. Astra costs $8 per million input tokens and $40 per million output, 20 percent below its predecessor, while Sol keeps GPT-6.1 Sol's $2 input and $10 output pricing. OpenAI claims Astra leads Agents' Last Exam and roughly doubles Claude Fable 5.2's pass rate on Harvey LAB-AA at about one-sixth the cost, but Artificial Analysis reports coding scores flatlined and Epoch AI found Astra barely improved over GPT-6 Astra on real-world research tasks.
- Hacker News · AIGPT-6.2 Sol moves the Pareto frontier in Artificial Analysis score vs. cost
Artificial Analysis places GPT-6.2 Sol on the Pareto line of its Intelligence Index versus weighted average cost per task, marking it as sitting in the most attractive quadrant of that chart. The Intelligence Index v5.0.1 aggregates 9 evaluations, including AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode-AA 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset) and Terminal-Bench-Science 0.1. The page also breaks out capability indexes for finance, legal, engineering and other domains, plus cost, speed, token use and context window metrics.
- OpenAI NewsPicksOpenAI releases GPT-6.2 Astra and Sol
OpenAI released GPT-6.2 Astra and Sol, which it says deliver better professional outputs, coding, computer use and efficiency at lower cost. The models are available across ChatGPT Work, Codex and the API.
- Artificial Analysis ArticlesPicksOpenAI releases GPT-6.2 Astra and GPT-6.2 Sol, taking #2 on Artificial Analysis Intelligence Index
OpenAI's GPT-6.2 Astra and GPT-6.2 Sol rank #2 on the Artificial Analysis Intelligence Index, trailing only Anthropic's Claude Fable 5.2. Astra gains 5 points over GPT-6 Astra, overtaking Gemini 4 Carbon (60) and Claude Opus 5.5 (59) at top effort settings, while Sol gains 2 points to 57 at roughly 15% of the Cost per Task. Astra is priced at $8/$40 per million tokens, 20% below GPT-6 Astra, with cached input dropping to $0.40 from $1.00; Sol stays at $2/$10. Neither model improves the Coding Agent Index, where Astra is flat and Sol loses 3 points, driven by a 14-point drop in SWE-Atlas-QnA-AA.
Heat for this event
- Current, comparable range
- 49
- Peak, comparable range
- 51November 18 07:00
- Change in 24 hours
- –
The trend compares the same participants over hours observed continuously, so its range can be narrower than the current heat count. Hover or tap the chart for hourly heat; the arrow keys move through it.