OpenAI ships GPT-6.2 Astra and Sol, and Sam Altman singles out Sol's Ultrafast mode
The two models are live in ChatGPT Work, Codex and the API. Astra's API price drops 20%, Sol keeps GPT-6.1 Sol's rates, and Ultrafast reaches Pro 200 subscribers for the first time, starting with Sol.
OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on Tuesday. Chief executive Sam Altman wrote on X that “6.2 Sol and Astra are out”, calling the models “very token-efficient” and adding that “6.2 Sol ultrafast is something I think people will really enjoy.”
Two models, two jobs
In its launch post, OpenAI describes Astra as its new frontier model “for your hardest projects” and Sol as an “efficient, versatile model for everyday work”. The company says Astra’s slides, spreadsheets and documents follow templates and professional conventions more closely, and that it writes cleaner code and better frontends for websites, games and 3D design. The Verge notes the same pitch, that Astra’s office documents “read less like AI”.
Background
Earlier stories on this thread.
What happened next
Later stories on this thread.
OpenAI says Sol is smarter and more efficient than GPT-6.1 Sol and lets paid users finish more tasks within the same plan. It adds that many of its own staff use Sol as their default model.
Where you can use them
According to the developer announcement, both models are rolling out in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users, and in the API as gpt-6.2-astra and gpt-6.2-sol. Enterprise and Edu admins have to switch the models on first. Pro 500 subscribers can use Ultrafast with either model, and OpenAI is bringing Ultrafast to Pro 200 subscribers for the first time, starting with Sol. Ultrafast is available for both models in the API from launch.
Pricing
OpenAI lists Astra at $8 per million input tokens, $0.40 for cached input and $40 for output, which it says is 20% lower than GPT-6 Astra on input and output with a 60% discount on cached input. Sol keeps GPT-6.1 Sol’s rates of $2 for input, $0.10 for cached input and $10 for output.
The benchmark claims
OpenAI says Astra scores 5.2 points higher than GPT-6 Astra on Agents’ Last Exam, a test of multistep professional work, at less than a third of the cost, and that on Cognition’s FrontierCode 2.0 Sol comes within 0.5 points of Astra at about one fifth of the cost. These are the company’s own figures. It also says it is launching both models with tighter safeguards for higher-risk cybersecurity help because of their stronger cyber abilities.
Our read
Analysis: the headline here is price as much as capability. Holding Sol’s price while claiming better results, and cutting Astra’s, puts pressure on rivals’ everyday-model pricing. Independent benchmark results will show whether the efficiency claims hold outside OpenAI’s own tests.
This report is not legal, financial, or health advice. See the disclaimer.
Primary sources
- Sam Altman on X: 6.2 Sol and Astra are out Sam Altman on X
- Introducing GPT-6.2 Astra and Sol OpenAI
- GPT-6.2 Astra and Sol are now available OpenAI Developer Community
- OpenAI releases its updated GPT-6.2 Astra model The Verge