OpenAI releases GPT-5.6 with three variants after government security review
OpenAI is releasing GPT-5.6 to the public on July 9 following government security review under a Trump administration AI cybersecurity order. The release includes three variants: Sol (strongest), Terra (GPT-5.5 performance at half the cost), and Luna (lowest cost option).
GPT-5.6 Sol — Quick Specs
OpenAI releases GPT-5.6 with three variants after government security review
OpenAI will publicly launch all three GPT-5.6 variants — Sol, Luna, and Terra — on July 9, following approval from the Trump administration after a mandatory security review process.
The company initially released GPT-5.6 to a "small group of trusted partners" in late June under a June 2026 presidential AI cybersecurity order requiring companies to submit their most powerful models for government review 30 days before public release. OpenAI submitted the models voluntarily and received expedited approval in less than 30 days.
According to Axios, the Department of Commerce's Center for AI Standards and Innovation conducted additional testing, with OpenAI sending technical experts to Washington, DC to address questions and concerns in real time.
Three variants at different price points
GPT-5.6 comes in three configurations:
- Sol: OpenAI's strongest model yet at $5 per million input tokens and $30 per million output tokens
- Terra: Designed for everyday use, claims similar performance to GPT-5.5 at half the cost — $2.50 per million input tokens and $15 per million output tokens
- Luna: Lowest-cost option at $1 per million input tokens and $6 per million output tokens
OpenAI stated it was "expanding preview access globally now" via X (formerly Twitter). The company previously said it "doesn't believe this kind of government access process should become the long-term default" but complied to ensure a timely public release.
Industry-wide compliance
Anthropic also underwent similar government review for its Mythos cybersecurity and Fable models, blocking all access including to foreign nationals during the review period. The company has since received permission to redeploy the Mythos 5 model and will eventually release Fable 5.
What this means
The expedited review process suggests the government is developing faster protocols for AI model approval, though the precedent of mandatory pre-release reviews for frontier models remains controversial. The pricing structure shows OpenAI is segmenting capabilities across price tiers more aggressively than previous releases, with Terra positioned as a cost-optimized alternative to GPT-5.5 rather than a capability upgrade. Whether Sol represents a significant architectural advance or incremental improvement over GPT-5.5 remains to be demonstrated through independent benchmarks.
Related Articles
Simon Willison's Pelican Benchmark Shows GPT-6 Astra Outperforming GPT-5.6 Sol at Every Reasoning Level
Developer Simon Willison ran his signature 'pelican riding a bicycle' SVG test on newly-accessed GPT-6 Astra across five reasoning levels, comparing results against GPT-5.6 Sol, Terra, and Luna. Even Astra's lowest reasoning setting reportedly beat every Sol output, though Astra costs roughly twice as much per token.
OpenAI Lists GPT-6 Astra Pro on OpenRouter: Same Model, Higher-Compute Reasoning Mode
GPT-6 Astra Pro, now listed on OpenRouter, is the existing GPT-6 Astra model configured to run with reasoning.mode set to 'pro' for higher-quality output on complex tasks. It carries a 1M-token context window and tiered pricing from $5/$25 to $20/$100 per million input/output tokens depending on the serving tier.
OpenAI Releases GPT-6 Astra, First Model to Cross 'Critical' Cybersecurity Threshold
OpenAI has begun rolling out GPT-6 Astra, the first model to reach the company's internal 'Critical' cybersecurity threshold. Access is being phased, with companies in OpenAI's Daybreak cybersecurity program getting priority following added safeguards after a prior model containment breach.
OpenAI's GPT-6 Astra Reportedly Automates AI Engineering Tasks at Under $6 an Hour, According to Latent Space Testing
A Latent Space report describes GPT-6 Astra, a new OpenAI model the blog says can autonomously handle AI engineering tasks—training models, labeling data, deploying systems—at an estimated cost of under $6 per hour. The claims, including 97.6% on FrontierMath and 99.9% on ARC-AGI-3, come from independent blog testing rather than an official OpenAI announcement.
Comments
Loading...