GPT-6 Luna
The fastest and most cost-efficient model in the GPT-6 generation
free access available

- Context
- 1,050,000 tokens
- Output
- up to 128,000 tokens
- Input
- Text, images
- Knowledge
- up to May 2026
- Developer
- OpenAI
- Access
- ChatPlus PRO
GPT-6 Luna is the entry-level model in the GPT-6 family, introduced by OpenAI on September 22, 2026. It is built for clear everyday tasks: summarizing a document, extracting data, quickly answering a question, or analyzing correspondence. Luna is half the price of GPT-5.6 Luna, has the freshest knowledge in the lineup — up to May 2026 — and the same context window of up to 1.05 million tokens. Available on ChatPlus with a PRO subscription and does not use allowance.
Luna is a model for tasks with a clear goal: summarize, extract, organize, answer quickly. That is exactly how OpenAI describes it, and there is no false modesty in that — these are the tasks most people have most often.
Compared with GPT-5.6 Luna, the new version is half the price for input and more than half the price for output, while matching it in independent intelligence tests. One interesting detail: the entry-level model has the freshest knowledge in the family — up to May 2026 — while Sol and Astra stop at April.
Luna is not the best choice for complex multi-file code — switch to GPT-6 Sol in the same conversation instead. But for correspondence, notes, and quick lookups, it is more than capable. On ChatPlus, GPT-6 Luna is available with a PRO subscription and does not use allowance.
What it's good at
Speed
Responds quickly — ideal for short questions and high-volume work.
Summarization and extraction
Condenses long documents and extracts numbers, dates, and names in the format you need.
Fresh knowledge
Trained on data through May 2026 — more recent than Sol and Astra.
No allowance usage
Does not use your 5-hour limit on ChatPlus, so you can use it freely.

GPT-6 Luna and every other model — one plan, one window.
Where it's strongest
GPT next to the others
How GPT-6 Luna and the flagships of other labs look side by side. They're all in one ChatPlus subscription.
| Claude | DeepSeek | GPT | Gemini | Grok | GLM | Kimi | Qwen | Muse | GLM 5.3 Flash | |
|---|---|---|---|---|---|---|---|---|---|---|
| Developer | Anthropic | DeepSeek | OpenAI | xAI | Z.ai | Moonshot | Alibaba | Meta | Z.ai | |
| Model | Sonnet 4.6 | DeepSeek V4.1 Reasoner | GPT-6 Luna | Gemini 4 Argon | Grok 4 | GLM 5 | Kimi K3 | Qwen 3.8 Max | Muse Spark 1.3 | GLM 5.3 Flash |
| Context | 200K | 128K | 1M | 1M | 256K | 1M | 256K | 256K | 1M | 1M |
| Understands images | — | — | ||||||||
| Live web data | — | — | — | — | — | — | — | |||
| Open source | — | — | — | — | — | |||||
| Key strength | Text and code | Reasoning | Speed | Multimodality | Timeliness | Code and agents | Long context | Complex tasks | Agents and code | Code and agents |
| On ChatPlus |