The short answer
GPT-6 Sol vs GPT-6 Luna: specs side by side
| GPT-6 Sol | GPT-6 Luna | |
|---|---|---|
| Made by | OpenAI | OpenAI |
| Released | September 22, 2026 | September 22, 2026 |
| API model ID | gpt-6-sol | gpt-6-luna |
| Context window | 1.05M tokens (922K input) | 1.05M tokens (922K input) |
| Max output | 128,000 tokens | 128,000 tokens |
| Knowledge cutoff | April 20, 2026 | May 18, 2026 |
| Inputs | Text and images | Text and images |
| API price | $2 input / $10 output per 1M tokens | $0.10 input / $0.50 output per 1M tokens |
| In AimiChat | Premium quality · Aimi Ultra ($50/month) | Fast quality · Free ($0) |
Which to use for each task
Quick questions and rewrites
Luna is built for speed and is included free in AimiChat.
Hard analysis and planning
Sol is built for demanding reasoning where Luna may miss a step.
Coding beyond small snippets
Sol scores higher on OpenAI's reported coding benchmark (68.8% vs 66.6% on DeepSWE v1.1 at max effort) and handles longer tasks better.
Summaries and extraction
Luna is designed for exactly this kind of repeatable work, at a fraction of the cost.
What to know before you choose
On OpenAI's reported DeepSWE v1.1 results the gap is small (68.8% for Sol vs 66.6% for Luna, both at max effort), which is why Luna is good enough for many everyday coding questions.
In AimiChat, GPT-6 Luna is on the Free plan and GPT-6 Sol is available from the Aimi Ultra plan.
Published benchmarks
Each company reports different tests with different settings, so these are listed per model rather than head to head.
| Model | Benchmark | Score | Reported by |
|---|---|---|---|
| GPT-6 Sol | DeepSWE v1.1 (max effort) | 68.8% | OpenAI launch post, as reported by VentureBeat and Codersera |
| GPT-6 Sol | OSWorld 2.0 offline (xhigh effort) | 60.5% | OpenAI launch post, as reported by VentureBeat and Codersera |
| GPT-6 Luna | DeepSWE v1.1 (max effort) | 66.6% | OpenAI launch post, as reported by VentureBeat and Codersera |
Try both in AimiChat
GPT-6 Sol runs on the Premium quality and GPT-6 Luna on Fast. On a plan that includes both, you can send the same prompt twice in one chat, switching quality in between, and compare the answers directly. Read more about GPT-6 Sol and GPT-6 Luna, or see which plan includes which model.
Sources
Specifications and prices come from the model makers and API listings below. Benchmark figures are reported by the company that made the model and are not independent tests.