AI News
Written by Joshua Reiher
By Joshua Reiher
Some links on this page are affiliate links. If you sign up through them we may earn a commission, at no extra cost to you.
![]()
OpenAI has released two new models: GPT-6.2 Astra, which it calls its "new frontier model for your hardest projects", and GPT-6.2 Sol, an "efficient, versatile model for everyday work". Both were announced on 17 November 2026 (US time) and are rolling out in ChatGPT Work and Codex, and in the OpenAI API.
OpenAI says both models get better results on demanding tasks while using fewer output tokens, which means you can get more done before you hit your plan's usage limits.

Image: OpenAI
Astra replaces GPT-6 Astra as OpenAI's top model. OpenAI says its slides, spreadsheets and documents "read less like AI" and follow the templates you give it, and the conventions of fields like law, sales, finance and accounting, more closely. It can match a slide template's layout, colours and fonts from an original file or even a screenshot.
OpenAI also says Astra writes cleaner code, builds better-looking websites and 3D experiences, and is more precise when it controls a computer, down to individual parts in CAD and simulation tools. OpenAI's developer guide adds that Astra is more concise and direct by default, and may suggest a next step and wait for you to confirm when a task is unclear.
Sol is the cheaper, faster option and replaces GPT-6.1 Sol. OpenAI says it is smarter and more efficient than GPT-6.1 Sol and lets you complete more tasks on the same paid plan. OpenAI also pitches it for everyday tasks outside work, such as comparing homes or planning meals.
According to OpenAI's help centre, both models are rolling out in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, Edu, ChatGPT for Healthcare and FedRAMP workspaces. Two things are worth knowing:
Availability also depends on your workspace settings and how far the rollout has reached. OpenAI says to update the ChatGPT desktop app and the Codex CLI if you do not see the new models.
Developers can use both models in the API as gpt-6.2-astra and gpt-6.2-sol, with a context window of 1,050,000 tokens and up to 128,000 output tokens.
For ChatGPT, OpenAI's pricing page lists Plus at US$20 a month and Pro from US$100 a month (prices in US dollars). The new models use your plan's included Work and Codex allowance.
In the API, OpenAI's official prices per million tokens are:
Requests over 272,000 input tokens cost more. A faster "Ultrafast" mode is charged at 3x the standard rate until 28 February 2027 and 6x from 1 March 2027, so the 3x rate is a launch price, not the normal price.
These are OpenAI's own published results:
OpenAI says both models meet the "Critical" threshold for cybersecurity under its Preparedness Framework, and they launch with tighter safeguards for higher-risk security help. OpenAI notes this may also block some legitimate security research.
If you pay for ChatGPT Plus or Pro and use Work or Codex for documents, slides, spreadsheets or coding, you should see the new models in the Work model picker soon. If you mostly use regular Chat, nothing changes for now. Developers get a stronger top model for less money, and Sol is worth testing as a cheaper default.
We checked every claim on 18 November 2026 against OpenAI's own pages:
Already a member? The same buttons sign you in.
We'll email you a 6-digit code. No password needed. By continuing you agree to our Privacy Policy.
We sent a 6-digit code to . It expires in 10 minutes.
No email? Check your spam folder, or ·