AIToday
Large Language ModelsAI Business & IndustryThe Verge AIPublished: Jun 27, 2026, 04:00 JST

OpenAI launches GPT-5.6 suite under Trump admin oversight

OpenAI launches GPT-5.6 suite under Trump admin oversight

3 Key Points

  1. What happened

    OpenAI unveiled GPT-5.6, a model suite with three tiers—Sol (flagship), Terra (medium-tier), and Luna (affordable)—priced at $5 input / $30 output per million tokens for Sol, roughly half the cost of Anthropic's Claude Fable 5. The preview is being monitored by the Trump administration, which will approve customers on a case-by-case basis.

  2. Why it matters

    The Trump administration requested OpenAI stagger the release for security review, reflecting concerns about dual-use AI capabilities like cybersecurity. OpenAI designed the preview period to test whether safeguards occasionally block legitimate work, particularly in areas where defensive and offensive activity look similar. The company says it cooperated with the US government but hopes this case-by-case approval does not become the long-term default for future releases.

  3. What to watch

    OpenAI says the model suite should be generally available in the coming weeks. Sol has the company's 'most robust safety stack to date' and dedicates approximately 700,000 A100e GPU hours to automated red-teaming, with third-party testers continuing to evaluate it for the next two weeks.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

FAQ
How much does GPT-5.6 Sol cost?
Per million tokens, GPT-5.6 Sol costs $5 input / $30 output, roughly half the cost of Anthropic's Claude Fable 5, which is $10 input / $50 output. Terra costs half the price of Sol, and Luna costs less than half the price of Terra.
When will GPT-5.6 be available to everyone?
OpenAI said the model suite should be generally available in the coming weeks. During the preview period, the Trump administration approves customers on a case-by-case basis.
What makes GPT-5.6 Sol different from earlier versions?
OpenAI says Sol has the company's 'most robust safety stack to date' and is trained to refuse prohibited cyber assistance, including when users attempt to disguise their intent or jailbreak the model. Sol also has two additional modes: a 'max' mode for deeper reasoning and an 'ultra' mode for leveraging sub-agents.

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleU.S. government approval now gates AI model releases for OpenAI, Anthropic