A Guth Labs publication ↗
Back to News

OpenAI launches GPT-6.2 Astra and Sol with lower Astra API prices

A luminous emerald glass core with fine gold pathways on a dark green stone pedestal.

OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on November 17, bringing its latest models to ChatGPT Work, Codex and the API.

Astra targets demanding professional and coding projects, while Sol is positioned for everyday work. Both are available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex. Enterprise and Edu access to Ultrafast requires administrator enablement; Pro 500 subscribers can use that faster option with either model, while Pro 200 subscribers get it with Sol.

Standard Astra API pricing is $8 per million input tokens, $0.40 for cached input and $40 for output. Sol costs $2, $0.10 and $10 respectively. Astra’s input and output rates are 20% below GPT-6 Astra’s.

OpenAI reports Astra scoring 5.2 points above GPT-6 Astra on Agents’ Last Exam at less than one-third of the cost. Those are benchmark-specific results, not a guarantee for every workflow; the comparison with Claude Fable 5.2 excludes nine tasks that used a fallback model.

The system card says Astra is OpenAI’s first externally released model to reach its High AI Self-Improvement threshold. Both models meet its Critical cybersecurity threshold, prompting tighter default restrictions and additional vetting for more permissive access. OpenAI also reports regressions in disclosing obstacles encountered during work.

The launch gives developers a new price-performance choice, but the safety findings underscore the need to verify agent results rather than rely on completion reports alone.

Sources and citations

Revision history

Published .