
OpenAI is upgrading both of its flagship AI models, launching GPT-6.2 Astra and GPT-6.2 Sol in ChatGPT Work and its Codex coding tool, according to 9to5Mac. The headline improvement is efficiency. OpenAI says both models can handle demanding tasks using fewer output tokens, so users can get more done before they hit their usage limits. The release comes just a couple of months into the GPT-6 era, which started with Astra in September and continued with Sol's 6.1 update at OpenAI's DevDay a few weeks later. "Some of you have been asking very politely. Some of you have just been asking," OpenAI posted on X when it announced the launch.
Astra is still the model OpenAI recommends for the toughest jobs. The company says it can now take a screenshot of a single slide and use it as a template for an entire presentation, matching the layout, colors and even font sizes. In Codex, OpenAI says Astra is better at changing and testing code inside existing projects, and early testers reported cleaner code and fewer rounds of revisions. Sol is the cheaper everyday option for coding and writing, and it comes surprisingly close to Astra. On Cognition's FrontierCode 2.0 benchmark, which checks whether code changes are good enough to merge into real projects, OpenAI says Sol finished within half a point of Astra at roughly one-fifth of the cost.
Both models are available starting today for Plus, Pro, Enterprise and Edu users. The $200-a-month Pro tier gets OpenAI's Ultrafast mode on Sol for the first time, while the $500-a-month tier gets it on both models, as do eligible Enterprise and Edu workspaces once an admin turns it on. OpenAI says Ultrafast can generate responses up to 4.5 times faster than standard speed, though it burns through usage allowances more quickly. Developers can also use both models through the API today. Astra's standard pricing is 20% lower than GPT-6 Astra's, and Sol keeps the same pricing as GPT-6.1.

