GPT-5.6 Goes Public: OpenAI Launches Sol, Terra, and Luna With ChatGPT Work

OpenAI released its GPT-5.6 model family on July 9 after a government security review, debuting three tiers — Sol, Terra, and Luna — alongside ChatGPT Work for enterprise teams. The flagship Sol model claims the top spot on the Coding Agent Index, overtaking Anthropic's Fable 5.

OpenAI officially released the GPT-5.6 family of models to the public on Thursday, July 9, following a limited preview and a government cybersecurity review that had briefly restricted wider access. The launch introduces three variants: Sol, the flagship workhorse; Terra, a balanced intermediate option; and Luna, the budget-friendly entry point.

Sol immediately claimed the top position on the Artificial Analysis Coding Agent Index with a score of 80 — 2.8 points above Anthropic's recently released Fable 5 — while using less than half the output tokens. Terra performs just above Fable 5, and Luna outperforms Opus 4.8, giving OpenAI a clean sweep across the pricing ladder. CEO Sam Altman told CNBC that Sol is 54% more token efficient for coding tasks compared to previous models.

The company also describes GPT-5.6 as its "strongest cybersecurity model yet," with defensive capabilities including threat modeling, code review, patching, and blue teaming. Those cybersecurity chops were precisely what drew scrutiny: the Trump administration had requested a 30-day review window before public release, but ultimately gave OpenAI the green light after additional testing.

Alongside the models, OpenAI launched ChatGPT Work — a desktop, web, and mobile workspace that combines ChatGPT with Codex, letting non-coders create documents, spreadsheets, presentations, and websites from a single interface. The move positions OpenAI directly against Anthropic's enterprise-focused strategy.

Pricing starts at $1 per million input tokens and $6 per million output tokens for Luna, scaling to $5 input / $30 output for Sol. GPT-5.6 is available now through ChatGPT, Codex, and the OpenAI API.

Sources: TechCrunch, Engadget, Dataconomy