GPT-5.6 API
3 models · OpenAI-compatible · max reasoning effort · ultra mode with subagents
GPT-5.6 API — Early Access
OpenAI's next-generation family — Sol, Terra, and Luna — was previewed on June 26, 2026. Public API access is rolling out in the coming weeks. Join the EvoLink waitlist to get access on day one.
We'll only email you about GPT-5.6 availability. No spam.
The GPT-5.6 family
Three models, one API — pick the tier that fits your workload.
| Model | Best for | Price / 1M (in / out)Preview pricing |
|---|---|---|
| Sol | Flagship — the strongest, deepest-reasoning model, with a new max reasoning effort and ultra mode | $5 / $30 |
| Terra | Balanced everyday work — GPT-5.5-class performance at roughly half the cost | $2.50 / $15 |
| Luna | Fast and affordable — high-volume, latency-sensitive, or budget-conscious workloads | $1 / $6 |
Model lineup and pricing per OpenAI's preview announcement (June 26, 2026). Context window, max output, and final model IDs are being confirmed and may change at general availability.
See the full GPT family →Need it today? Build on GPT-5.5
The GPT-5.6 API isn't open to everyone yet. GPT-5.5 is available right now on EvoLink — 1M context window, $5 / $30 per 1M tokens — and moving to GPT-5.6 later is a one-line model-ID change on the same OpenAI-compatible endpoint.
GPT-5.6 next steps
Build on GPT-5.5 today, grab an API key, and explore the OpenAI model family on EvoLink.
GPT-5.6 Early Access - FAQ
When will the GPT-5.6 API be available?
OpenAI previewed GPT-5.6 (Sol, Terra, and Luna) on June 26, 2026 as a limited preview, with general availability planned in the coming weeks. Join the waitlist and we'll email you the moment it's live on EvoLink.
What is the GPT-5.6 family?
It's three models on one API: Sol is the flagship with the strongest reasoning, Terra is the balanced everyday model (GPT-5.5-class performance at about half the cost), and Luna is the fast, low-cost tier for high-volume workloads.
How much will GPT-5.6 cost?
Per OpenAI's preview: Sol is $5 / $30, Terra is $2.50 / $15, and Luna is $1 / $6 per 1M input / output tokens. Final EvoLink rates are confirmed when the API launches.
What's new versus GPT-5.5?
GPT-5.6 adds a new max reasoning effort tier and an ultra mode that orchestrates subagents for complex work. Per OpenAI's preview it reaches state-of-the-art results on Terminal-Bench 2.1 and cybersecurity benchmarks, ahead of GPT-5.5.
What do I get by joining the waitlist?
Early notification and day-one access on EvoLink when the GPT-5.6 API opens — one API key and an OpenAI-compatible endpoint, so you can switch with a single model-ID change.
Can I start building now?
Yes. Build against GPT-5.5 on EvoLink today using the standard Chat Completions / Responses pattern; upgrading to GPT-5.6 later is a model-ID change, not a rewrite.
Other GPT models on EvoLink
Learn more about GPT-5.6
GPT-5 API pricing guide 2026
Base rates, cached input & cost tips for every GPT-5 model.
GPT-5.4 vs Claude Opus 4.6 vs Gemini 3.1 Pro
Three-way flagship comparison with pricing & benchmarks.
Best LLM for coding agents
Compare LLMs on coding-agent cost & reliability.
All AI models
Browse every model on EvoLink
GPT-5.6 lineup, pricing, and performance figures are from OpenAI's preview (June 26, 2026) and may change at general availability.