GPT-6.2 Astra and Sol Launch in ChatGPT Work, Codex, and API
YYSI publication · Source announcement
What changed
OpenAI says GPT-6.2 ships as two models, Astra and Sol, available in ChatGPT Work, Codex, and the Responses API. Astra is positioned for complex reasoning, while Sol is aimed at everyday coding and writing.
Who it affects
Developers can call gpt-6.2-astra or gpt-6.2-sol in the Responses API, and both support Ultrafast at launch. Sol is described as completing more tasks on the same paid plan than GPT-6.1 Sol.
How to try it
Confirm your plan or workspace grants access to the chosen model, since Enterprise and Edu admins must first enable the models in workspace settings.
- Choose one representative task and record the current model or integration as a baseline.
- Check the announcement for the exact product, access tier and entry point. Run the same small input through the new option if your account has access.
- Compare output quality, latency and reported usage before changing your default. Apply verified rates to your own token counts in the cost calculator.
What it costs
OpenAI’s launch announcement lists standard API rates in USD per million tokens. Astra is $8 for input, $0.40 for cached input and $40 for output. Sol is $2 for input, $0.10 for cached input and $10 for output.
Official announcement ↗ · Read the news brief
AI-assisted explainer based on the official announcement. The steps are a suggested evaluation, not YYSI test results. Editorial standards