GPT-6 Luna
GPT‑6 Luna is a more affordable GPT‑6 model that incorporates alignment and capability improvements from Astra to deliver better factuality and coding/agent performance than its GPT‑5.6 predecessor at much lower cost. It is available in the OpenAI API (gpt-6-luna) and in certain ChatGPT tiers, and is accessible to free/Go users on the desktop app for Luna specifically.
What it does
Developers can use it through the API for coding tasks in a range of programming languages at a lower estimated evaluation cost than GPT-5.6 Sol.
- Availability
- Try our latest models through the API
- Pricing model
- API cost
- Deployment
- OpenAI API, ChatGPT Work, Codex, and the desktop app.
- Privacy
- Not established
- Licence
- Not established
Capability
Modality
Access
Deployment
Pricing
Target user
Evidence
14 claims- ChatGPT can create tools such as calculators, bill splitters, and games in a conversation.
- The rollout starts globally for Plus, Pro, Business, and Enterprise tiers, with Free and Go to follow tomorrow.
- The update applies to the Chat experience in ChatGPT.
- Responses can include graphics, tappable buttons, forms, charts, and interactive experiences.
- Enterprise availability depends on workplace admin settings.
- GPT-6 Luna at max effort exceeds GPT-5.6 Sol at medium effort on OSWorld 2.0 offline. [2]
On **OSWorld 2.0 offline**, GPT‑6 Sol at xhigh effort achieves a similar score to Claude Opus 5 at medium effort—60.5% versus 60.3%—at approximately 80% lower cost per task.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna is available in the OpenAI API as gpt-6-luna. [2] [3]
In the OpenAI API, they are available as `gpt-6-sol` and `gpt-6-luna`.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - Free and Go users can access GPT-6 Luna in the desktop app. [2]
Free and Go users can access GPT‑6 Luna in the desktop app.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna API pricing is $0.10 per 1 million input tokens and $0.50 per 1 million output tokens. [2] [3] [5]
- GPT-6 Luna at max effort scores 66.6% on DeepSWE v1.1. [2]
GPT‑6 Luna at max effort scores 66.6%, comparable to Claude Opus 5 and Fable 5 at medium effort.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna reaches an average test pass rate of 40.3% across 20 languages at maximum reasoning effort. [4]
At maximum reasoning effort, GPT‑6 Astra achieves an average test pass rate of 87.4% across the 20 languages, up from 26.8% for GPT‑5.6 Sol.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna improves on GPT-5.6 Sol in every language tested. [4]
GPT‑6 Luna offers a lower-cost option that also improves on GPT‑5.6 Sol in every language tested.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna's estimated average API cost is $0.77 per sample, compared with $3.48 for GPT-5.6 Sol. [4]
These improvements come at an estimated average API cost of $0.77 per sample, compared with $3.48 for GPT‑5.6 Sol, approximately 78% lower.
OpenAI. Quoted text is outside CC BY 4.0 and keeps its original rights. - GPT-6 Luna is available through the API.
Sources
5
1
Verified
2
OpenAIOwner-controlled
https://openai.com/index/introducing-gpt-6-sol-and-luna
Retrieved 2026-10-28 00:27 UTC
Verified
3
OpenAIOwner-controlled
https://developers.openai.com/api/docs/changelog#2026-09-22
Retrieved 2026-10-28 00:27 UTC
Verified
4
OpenAIOwner-controlled
https://openai.com/index/low-resource-programming-languages
Retrieved 2026-10-28 00:27 UTC
Verified
5
MicrosoftReputable independent
https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/
Retrieved 2026-10-28 00:27 UTC
Verified
Versions
39| Version | Effective | Change | Changed fields | State |
|---|---|---|---|---|
| 39 | 2026-10-28 00:27 UTC | material change | capability, access, pricing, availability, pricing model | Current |
| 38 | 2026-10-28 00:27 UTC | material change | availability, pricing model, deployment | Current |
| 37 | 2026-10-26 01:36 UTC | evidence update | score | Current |
| 36 | 2026-10-26 01:36 UTC | material change | capability, access, pricing, availability, pricing model, deployment | Current |
| 35 | 2026-10-25 01:33 UTC | material change | capability, modality, access, pricing, availability, pricing model | Current |
| 34 | 2026-10-25 01:33 UTC | material change | capability, modality, availability, pricing model, deployment | Current |
| 33 | 2026-10-25 01:33 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 32 | 2026-10-24 02:04 UTC | material change | capability, modality, access, pricing, availability, pricing model | Current |
| 31 | 2026-10-24 02:04 UTC | material change | capability, access, pricing, availability, pricing model, deployment | Current |
| 30 | 2026-10-23 01:46 UTC | material change | capability, modality, access, pricing, availability, pricing model | Current |
| 29 | 2026-10-23 01:46 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 28 | 2026-10-22 01:53 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 27 | 2026-10-22 01:53 UTC | material change | capability, access, pricing, availability, pricing model, deployment | Current |
| 26 | 2026-10-21 02:48 UTC | material change | pricing, pricing model | Current |
| 25 | 2026-10-21 02:48 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 24 | 2026-10-21 02:48 UTC | material change | capability, access, pricing, availability, pricing model, deployment | Current |
| 23 | 2026-10-20 03:38 UTC | material change | capability, access, pricing, availability, pricing model | Current |
| 22 | 2026-10-20 03:38 UTC | material change | availability, pricing model, deployment | Current |
| 21 | 2026-10-19 04:38 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 20 | 2026-10-18 05:28 UTC | material change | capability, modality, access, pricing, availability, pricing model, deployment | Current |
| 19 | 2026-10-17 06:27 UTC | evidence update | score | Current |
| 18 | 2026-10-16 07:22 UTC | evidence update | score | Current |
| 17 | 2026-10-15 08:14 UTC | material change | availability, pricing model, deployment | Current |
| 16 | 2026-10-14 09:06 UTC | material change | availability, pricing model, deployment | Current |
| 15 | 2026-10-13 09:58 UTC | material change | availability, pricing model, deployment | Current |
| 14 | 2026-10-12 10:55 UTC | material change | availability, pricing model, deployment | Current |
| 13 | 2026-10-11 11:48 UTC | material change | availability, pricing model, deployment | Current |
| 12 | 2026-10-10 12:44 UTC | material change | availability, pricing model, deployment | Current |
| 11 | 2026-10-08 14:32 UTC | material change | availability, pricing model, deployment | Current |
| 10 | 2026-10-07 15:30 UTC | material change | availability, pricing model, deployment | Current |
| 9 | 2026-10-06 16:29 UTC | material change | availability, pricing model, deployment | Current |
| 8 | 2026-10-05 17:26 UTC | material change | availability, pricing model, deployment | Current |
| 7 | 2026-10-04 18:21 UTC | material change | availability, pricing model, deployment | Current |
| 6 | 2026-10-03 19:18 UTC | material change | availability, pricing model, deployment | Current |
| 5 | 2026-10-02 20:14 UTC | material change | availability, pricing model, deployment | Current |
| 4 | 2026-10-01 06:48 UTC | evidence update | pricing offers | Current |
| 3 | 2026-09-29 22:54 UTC | material change | capability, pricing model, deployment | Current |
| 2 | 2026-09-25 03:19 UTC | material change | capability, modality, pricing, availability, pricing model, deployment | Current |
| 1 | 2026-09-24 04:13 UTC | publication | None | Current |
In briefs
2- Backburner leads 25 AI developments in week 40, 2026 Weekly brief 2026-W40
- Perplexity introduces a decision model fine-tuned from Qwen3.8-27B Daily brief 2026-10-01
Original catalog content is licensed under CC BY 4.0, attributed to wowsuchai. Source quotations and linked content keep their original rights.
