OpenAI launches GPT-6.2 Astra and Sol AI models

Technology
November 19, 2026 · 8:00 AM3 min read

OpenAI released two new artificial intelligence models, GPT-6.2 Astra and GPT-6.2 Sol, on November 17, expanding its lineup with upgrades aimed at professional work, coding and computer use at a lower cost.

The models rolled out simultaneously in ChatGPT Work, the Codex coding tool and OpenAI’s developer API.

What the new models do

OpenAI introduced GPT-6.2 Astra as aimed at demanding professional work, while GPT-6.2 Sol is built for everyday use.

Starting the same day, GPT-6.2 Astra and Sol are available in ChatGPT Work and Codex for Plus, Pro, Enterprise, and Edu users.

On Agents’ Last Exam, a benchmark that tests multi-step professional workflows, GPT-6.2 Astra scores 5.2 points above its predecessor GPT-6 Astra at less than a third of the cost, according to OpenAI.

According to OpenAI’s announcement, API pricing lists GPT-6.2 Astra at $8.00 per million input tokens and $40.00 per million output tokens, while GPT-6.2 Sol costs $2.00 for input and $10.00 for output.

On the FrontierCode 2.0 coding benchmark built by Cognition, Sol nearly matches Astra on merge-ready changes to real codebases at about one-fifth the cost, OpenAI said.

Stronger safeguards for cyber capability

OpenAI said GPT-6.2 Astra is its most capable model ever for cyber, jumping 22 points over GPT-6 Astra on a benchmark it calls ExploitBench Internal Port.

Because of that jump in capability, OpenAI said it is launching GPT‑6.2 Astra and Sol with tighter safeguards for higher-risk cybersecurity assistance.

GPT-6.2 Astra is the first model OpenAI is deploying externally that it determined meets the High threshold for AI Self-Improvement under its Preparedness Framework , while GPT-6.2 Sol falls below it.

Separately, both models meet the Critical threshold for Cybersecurity capabilities and the High threshold for Biological and Chemical capabilities under OpenAI’s Preparedness Framework , according to the system card OpenAI published alongside the launch.

Part of a broader push for more capable, efficient models

The update follows a rapid cadence of releases this year. OpenAI had scrapped the planned October release of GPT-6.1 Astra after internal testing found the model did not meet the company’s safety and alignment standards , so GPT-6.2 Astra is the first upgrade to GPT-6 Astra since it began rolling out on Sept. 3 , while GPT-6.2 Sol succeeds GPT-6.1 Sol, which OpenAI upgraded Sol to GPT-6.1 a few weeks later at DevDay .

The new versions build directly on that work, with Astra’s standard input and output prices 20% lower than with GPT-6 Astra, while Sol keeps its GPT-6.1 pricing.

Independent coverage of the launch noted the practical upside for users managing usage limits. OpenAI says both can finish demanding tasks with fewer output tokens, so you can get more done before hitting your usage limit.

For businesses weighing AI tool access on company devices, that efficiency push mirrors a broader workplace trend GCN has tracked, where office laptops run many browser add-ons alongside AI assistants.

Why the release matters

The GPT-6.2 launch underscores how quickly leading AI labs are iterating on both raw capability and operating cost, a dynamic that shapes which companies and governments can afford frontier-level AI tools.

It also lands amid wider efforts to broaden AI literacy and access, including state-level training partnerships and research funding for AI PhD fellows.

Developers can access the new models directly through the OpenAI API as gpt-6.2-astra and gpt-6.2-sol, with an Ultrafast option available for both, according to OpenAI.

Hugo Rojas
Editor
Hugo Rojas is the editor of GCN. With a Master of Science in Engineering, he specializes in technology, data, and science, and brings a human-centered perspective informed by psychology.