AI Models & Companies · Major AI Developments Explained

What's new in OpenAI's GPT-5.6 model family (Sol, Terra, and Luna)?

GPT-5.6 is OpenAI's model family released July 9, 2026, split into three tiers — Sol (flagship, for the hardest problems), Terra (balanced, for high-volume business tasks), and Luna (fast and low-cost, for everyday work) — and it was the first frontier model to go through a formal US government review before wider release.

Key takeaways

GPT-5.6 splits into three tiers: Sol (flagship), Terra (balanced), and Luna (budget/fast).
It became generally available July 9, 2026, after a US government review process that initially limited access to around 20 vetted organizations.
OpenAI cut Luna's price by 80% and Terra's by 20% later in July 2026.
Each tier targets a different use case rather than being a single one-size-fits-all model.

Three Tiers, Three Different Jobs

GPT-5.6 isn’t a single model but a family of three: Sol, the flagship tier built for the hardest problems like complex coding and security research; Terra, a balanced tier aimed at high-volume business tasks such as customer support, internal tools, and document analysis; and Luna, the fastest and lowest-cost tier for everyday work like summarization, drafting, and routine automation.

An Unusually Formal Rollout

Before GPT-5.6 became generally available, OpenAI shared the models and release plans with the US government, and access was initially limited to roughly 20 vetted organizations while the Commerce Department’s Center for AI Standards and Innovation completed a review. Wider public access opened on July 9, 2026, once that review cleared — a more formal government-review step than has typically preceded prior model launches.

Pricing Moved Quickly After Launch

Later in July 2026, OpenAI cut the price of GPT-5.6 Luna by 80% and Terra by 20%, a substantial adjustment within weeks of general availability that meaningfully changed the cost calculus for high-volume use of the lower tiers.

Why the Tiered Approach Matters

Splitting a release into three explicitly different-purpose tiers, rather than one model with a single price point, reflects a broader industry pattern: matching model capability (and cost) to task complexity, rather than routing every request through the most expensive, most capable model available.

How This Compares to Prior OpenAI Releases

The three-tier structure itself isn’t new to GPT-5.6 — OpenAI has offered multiple capability tiers in prior model generations too — but the formal government review process before general availability was a notable departure from how earlier OpenAI model releases reached the public, marking a more structured, higher-scrutiny rollout than the company’s typical release pattern.

See the Full AI Model Release Timeline

Track every major model release from OpenAI, Anthropic, and Google since GPT-4 with our free AI Model Release Timeline — filterable by provider.

Go deeper

Frequently asked questions

Which GPT-5.6 tier should most people use?

Terra is positioned for typical high-volume business tasks like customer support, internal tools, and document analysis — a reasonable default for most everyday use, reserving Sol for genuinely hard problems (complex coding, security research) and Luna for simple, high-speed, low-cost tasks like summarization and drafting.

Why was GPT-5.6 reviewed by the US government before release?

OpenAI shared the models and release plans with the US government ahead of public launch, and the Commerce Department's Center for AI Standards and Innovation completed a review before clearing wider access — an unusual, formal step for a commercial model release, reflecting increased government scrutiny of frontier AI capability.

Sources

[1]GPT-5.6: Frontier intelligence that scales with your ambition — OpenAI

Written by Editorial Team

Last updated August 12, 2026

Get one well-sourced answer a week

No spam. Unsubscribe anytime.