← Latest briefing

Technology

OpenAI releases GPT-6.2 Astra and Sol models in ChatGPT Work and Codex

OpenAI says the models use fewer output tokens, and a faster Ultrafast mode is limited to higher-tier plans.

The short version

  • OpenAI released GPT-6.2 Astra and Sol in ChatGPT Work and Codex, with Astra recommended for the hardest projects.[9to5Mac · 9to5Google]
  • OpenAI says both models can finish demanding tasks with fewer output tokens, and Sol came within half a point of Astra at about one-fifth the cost.[9to5Mac]
  • OpenAI says Ultrafast mode generates tokens up to 4.5 times faster, available on the $200/month Pro plan with Sol and on both models for higher-tier plans with admin enablement.[9to5Mac]

Key facts

  • OpenAI released GPT-6.2 Astra and Sol in ChatGPT Work and Codex.[9to5Mac · 9to5Google]
  • OpenAI says both models can finish demanding tasks with fewer output tokens.[9to5Mac]
  • OpenAI says Sol came within half a point of Astra at about one-fifth the cost when both models were at maximum reasoning effort.[9to5Mac]
  • OpenAI says Ultrafast mode can generate tokens at up to 4.5 times the standard rate on either model.[9to5Mac]
  • Ultrafast is available with Sol on the $200/month Pro plan and on both models on the $500/month Pro plan and eligible Enterprise and Edu workspaces.[9to5Mac]
  • Astra's standard input and output prices are 20% lower than GPT-6 Astra, while Sol keeps GPT-6.1 pricing.[9to5Mac]
  • Astra is priced at $8 per million input tokens and $40 per million output tokens.[9to5Google]
  • Sol is priced at $2 per million input tokens and $10 per million output tokens.[9to5Google]

What remains uncertain

  • The performance, efficiency, and speed claims come from OpenAI and are not independently verified in the provided reporting.[9to5Mac]
  • The reporting says Ultrafast uses more of users' allowance, but how much more is not quantified.[9to5Mac]

Sources

Outlet counts describe coverage, not independent confirmation. Reports may share a wire service or original source.