GPT-6.2 Astra and Sol Come to GitHub Copilot: What They Do
💼 Business AI News

GPT-6.2 Astra and Sol Come to GitHub Copilot: What They Do

OpenAI's newest coding models are now available in GitHub Copilot, and here is what they are built for.

This article was written by AI. It passed automated fact and quality checks, and our editor spot-checks published articles every day.

See AI World HQ first on GoogleAdd us as a preferred source and our stories show up higher in Google’s Top Stories.Add on Google

If you write code, or work with people who do, you may have noticed AI tools creeping into every step of the process. Writing a function is only part of the job. Someone still has to check the work, and that is where these new models are aimed.

What the two models are

GitHub's announcement describes two models from OpenAI, both designed for coding work.

  • GPT-6.2 Astra is aimed at agentic coding and terminal work. "Agentic" means the AI can carry out a multi-step task on its own, such as reading files, making changes, and checking the results. "Terminal work" refers to the text-based command line where developers run programs.
  • GPT-6.2 Sol is aimed at complex agentic coding and in-depth code review. Code review is the process of reading someone else's code to catch bugs, risky patterns, and mistakes before they reach users.

OpenAI says both models show stronger coding judgment than earlier GPT-6 models. In plain terms, they are meant to make better decisions about the code in front of them.

What the early results say

These numbers come from the companies themselves, so treat them as their claims, not independent verdicts.

  • According to the announcement, both models caught significantly more real issues in code review than earlier GPT-6 models.
  • Both used fewer tokens to get there. A token is a small chunk of text that an AI reads and writes, and token counts often affect how much computing a task costs.
  • GitHub says Sol caught nearly twice as many real issues as GPT-6.1 Sol, while keeping its findings accurate.

Who can use them and where

According to GitHub, the models are available to users of Copilot Pro+, Max, Business, and Enterprise. You choose them in the model picker, the menu where you select which AI model a tool uses. GitHub lists many places where it appears, including Visual Studio Code, Visual Studio, the Copilot command-line tool, the Copilot coding agent, the GitHub Copilot app, github.com, GitHub Mobile on iOS and Android, JetBrains IDEs, Xcode, and Eclipse.

The rollout is gradual. If you do not see the models yet, GitHub says to check back soon.

Pricing follows usage-based billing at provider list prices. For the current rates, check the official "Models and pricing for GitHub Copilot" page instead of relying on any article, including this one.

Wrap-up

GPT-6.2 Astra and Sol are now available in GitHub Copilot for eligible plans, with Astra aimed at agentic coding and terminal work and Sol aimed at deeper code review. The companies report stronger results, but your own tests matter most. Your next step: open the model picker in the tool you use most, or ask your admin whether the models are enabled for your account, and try one small review task this week.

Keep reading

📬 The week’s AI, in your inbox

One friendly email every Sunday — the 5 stories that mattered, in plain English. No spam, unsubscribe anytime.

Was this helpful?

✦ Generated by AI in AI World HQ's automated newsroom, from official sources. Checked by automated fact and quality gates, and our editor spot-checks published articles every day. Spot a mistake? Use the buttons above.

← Back to all stories