GPT-5.6 Sol, Terra and Luna: tiers and pricing

GPT-5.6 Sol, Terra and Luna: tiers and pricing

GPT-5.6 is OpenAI's model generation from July 9, 2026, and it shipped as three models instead of one: Sol (the flagship), Terra (the balance of price and performance) and Luna (the fast, cheap one). All three share a 1.05 million token context window and the same knowledge cutoff — the difference is capability and price. Here's what sets them apart, what they cost, which one to pick, and what happened to the names when GPT-6 arrived.

When was GPT-5.6 released?

On July 9, 2026, OpenAI released GPT-5.6 — but not as a single model. The generation shipped in three flavors: Sol, Terra and Luna. It was a deliberate break from how OpenAI had named its models so far, and it says a lot about where the industry is heading.

The idea is simple but deliberate: the number (5.6) marks the generation, while the name marks a capability tier that can be updated on its own cadence. Sol can get a successor without Terra or Luna moving at all, and vice versa. That solves a problem many of us recognize — picking the "right" model has gotten harder with every release.

GPT-5.6 Sol, Terra and Luna — what sets them apart?

• GPT-5.6 Sol is the flagship. Built for heavy reasoning and long-horizon agentic work — the model OpenAI called its most capable to date at launch. Call the bare gpt-5.6 alias in the API and Sol is what you get.
• GPT-5.6 Terra is the balanced model. Performance on par with GPT-5.5, at half the cost — a natural default for everyday coding and agentic tasks.
• GPT-5.6 Luna is the fast, cheap variant. Built for smaller tasks at scale, where latency and price matter more than raw capability.

All three share the same technical foundation: a 1.05 million token context window, up to 128,000 tokens of output, and the same knowledge cutoff (February 16, 2026). You also pick a reasoning effort per request — from none to max — so the same model can be tuned for anything from quick answers to deep problem-solving.

Image: mikemacmarketing, Flickr (CC BY 2.0)

How much does GPT-5.6 cost?

Prices are in dollars per million tokens, input/output:

• Sol: $5 / $30 — unchanged since launch.
• Terra: $2.50 / $15 at launch, $2 / $12 after the cut.
• Luna: $1 / $6 at launch, $0.20 / $1.20 after the cut.

On July 30, 2026, just three weeks after launch, OpenAI cut Luna's price by 80% and Terra's by 20%, while Sol stayed where it was. A sign of how fast competition among the cheaper model tiers is pushing prices down.

Worth noting: send more than 272,000 tokens in a single request and a surcharge kicks in — double the input price and 1.5x the output price for the entire request, not just the overage.

On benchmarks like Terminal-Bench 2.1 (agentic coding), Sol edges out both Claude Mythos 5 and GPT-5.5 — narrowly, but measurably. Luna, the cheapest model in the family, still scores 62.7% on SWE-Bench Pro, which says a lot about how much capability now fits into the "fast and cheap" tier.

Which GPT-5.6 model does ChatGPT use?

In ChatGPT you don't pick a tier the way you do in the API. At launch, Luna became the default model for Free and Go users, while Sol powered both the instant answers and the thinking modes on paid plans. Terra was mainly an API option for people building their own solutions who want near-top capability at a lower price.

Sol, Terra or Luna — which should you choose?

• Pick Sol when the task is hard and has many steps: agentic work, debugging large codebases, analysis where a mistake costs more than the call.
• Pick Terra for everyday coding and most production workflows. It's where most people land once they've done the math.
• Pick Luna when you run large volumes of simple tasks — classification, summaries, document extraction — and latency and price matter most.

Whatever the tier: budget for the long-context surcharge if you're feeding large documents or codebases through the model, since it can triple your bill if you don't plan for it. And build with an abstraction layer. Whether you run OpenAI or Claude, a thin integration layer makes it easy to switch tiers — or providers — when the next model lands.

What the naming change actually means

By separating generation from capability class, OpenAI can update a single tier without forcing everyone to switch models at the same time. If you build a system around Terra, the idea is that the next "Terra" should feel like a natural upgrade — not a new name you have to research from scratch.

It echoes how Anthropic has split its own models into Opus, Sonnet and Haiku — just with a generation number layered on top. If you want to see how Claude's equivalent tiers, pricing and benchmarks stack up, we've put it together in our guide to Claude.

GPT-6 Sol and Luna — what changed?

The naming system only held for two months before it was reshuffled. On September 3, 2026, OpenAI released GPT-6 Astra, and on September 22 came GPT-6 Sol and GPT-6 Luna. But the names don't mean what they did in GPT-5.6:

• Astra is the new flagship, above everything else.
• Sol has moved down a step and is now the middle tier for everyday work — $2 / $10 per million tokens.
• Luna is still the fast, cheap tier — $0.10 / $0.50 per million tokens, compared with $0.20 / $1.20 for GPT-5.6 Luna.
• Terra doesn't exist in GPT-6, at least not yet.

So if you're searching for "GPT-6 Sol", it isn't the same model as GPT-5.6 Sol, and it plays a different role in the family. The number in front of the name is what counts. In ChatGPT, GPT-6 Sol and Luna are available on Plus, Pro, Business, Enterprise and Edu, and Luna also reaches Free and Go users in the desktop app. Update October 25, 2026: on October 7, GPT-6 replaced the old "Latest" option in Chat. On paid plans it now runs on GPT-6 Sol, while Free and Go users still get GPT-5.6 Luna according to OpenAI's help center. GPT-5.6 Sol can still be picked manually. In the API, Codex and ChatGPT Work, GPT-6 Sol has already been succeeded by GPT-6.1 Sol ($2 / $10 per million tokens).

The model race keeps going, and the names keep multiplying. What matters is understanding what each tier is actually built for — not just what it's called. If context windows and tokens are new to you, we cover the basics in what an LLM is.