Two models, two products at once

OpenAI announced new versions in its GPT-6 line. GPT-6.2 Astra and GPT-6.2 Sol are being rolled out together in ChatGPT Work and Codex. The company had started the GPT-6 generation with Astra in September and upgraded Sol to GPT-6.1 a few weeks later at DevDay.

The headline claim: fewer tokens

According to OpenAI, both models can finish demanding tasks with fewer output tokens. A token is a small chunk of text that language models process; fewer output tokens mean the same job is completed faster and at lower cost.

What it means for users

Spending fewer tokens is a practical advantage, especially on subscriptions with usage limits: users can complete more tasks before hitting the cap. That could matter for long-running work such as coding and workflow automation.

Limited detail

The announcement did not include benchmark results, pricing changes or the list of countries where the models are available. The picture will become clearer as those details emerge.