Model tracker / Decision models / OpenAI
OpenAI Decisions (GPT-6 Luna)
OpenAI Decisions (GPT-6 Luna) is a closed decision model from OpenAI: it answers typed questions, such as yes or no, pick one or pick several, with probabilities instead of writing text. It costs $0.10 per million input tokens (OpenAI API (public beta), OpenRouter, Vercel AI Gateway), and output is free. It scores 73.5 on the independent JevBench Capability Score (board v1.7.46), where Jev, the reference, scores 77.1.
- Made by
- OpenAI
- Weights
- Closed
- Size
- Undisclosed
- Where
- OpenAI API (public beta), OpenRouter, Vercel AI Gateway
- Images
- Yes
- Context
- Unpublished
- Fine-tuning
- Unknown
OpenAI Decisions (GPT-6 Luna) price
US dollars per million input tokens. Decision models don't write text, so none of the hosted ones charge for output.
| Where | Input | Output | Notes | Checked |
|---|---|---|---|---|
| OpenAI API (public beta), OpenRouter, Vercel AI Gateway | $0.10 | Free | Input only; no output or cache charges. Regional processing premiums and long-context multipliers apply. | 2026-10-14 |
OpenAI Decisions (GPT-6 Luna) benchmarks
| Benchmark | Result | What it measures | Reported by |
|---|---|---|---|
| JevBench v1.7.46 | 73.5 | Capability Score: accuracy and calibration on typed decisions, with Jev as the reference system | Independent |
| JevBench time | 0.30 s | Seconds per decision: measured end to end for hosted APIs, the board's estimate for self-hosted models | Independent |
| Reported speed | About 150 ms | Latency on the maker's own hardware | OpenAI |
Makers' speeds are measured on their own hardware and aren't comparable with each other or with JevBench's timings.
Decision models like OpenAI Decisions (GPT-6 Luna)
Every other decision model we track, by independent JevBench score.
| Model | Input / M | JevBench | Weights |
|---|---|---|---|
| OpenAI Decisions (GPT-6 Luna) | $0.10 | 73.5 | Closed |
| Jev 1.13 TypeSafe AI | $0.042 | 77.1 | Closed |
| Clef Cloudflare | $0.24 | 75.3 | Open |
| d1 Liquid AI | $0.04 | 74.4 | Closed |
| Microsoft-Decision-1 Microsoft | $0.042 | 70.8 | Closed |
| Clef-flash Cloudflare | $0.038 | 63.8 | Open |
| Clef-omni Cloudflare | $0.15 | 58.3 | Open |
| d1-3B Liquid AI | Self-host | 49.5 | Open |
| d1-omni-600M (experimental) Liquid AI | Self-host | 32.8 | Open |
| Laya Convai Innovations | Free | 32.5 | Open |
| CLM-8B Stanford and NVIDIA | Self-host | 20.1 | Open |
| Strands Decider 2B AWS Strands Labs | Self-host | – | Open |
Compare them all in the decision models table →
OpenAI Decisions (GPT-6 Luna) on dbggr
- Haiku 5.5 and four-cent decision models: the AI price war reaches the bottom tier Saturday edition, 2026-10-10
- Microsoft enters decision models at Jev's price as TypeSafe raises $870M Daily, 2026-10-09
- OpenAI publishes 722 AI-written math papers, claiming proofs of the quasi-Riemann hypothesis and Unique Games Daily, 2026-10-06
- Containing agents stopped being the labs' private problem Saturday edition, 2026-10-03
- Stop asking your LLM multiple-choice questions: a guide to decision models Article, 2026-10-02
- OpenAI launches Dots, agents that keep working after you log off Daily, 2026-09-29
OpenAI Decisions (GPT-6 Luna) FAQ
How much does OpenAI Decisions (GPT-6 Luna) cost?
OpenAI Decisions (GPT-6 Luna) costs $0.10 per million input tokens at OpenAI API (public beta), OpenRouter, Vercel AI Gateway. Input only; no output or cache charges. Regional processing premiums and long-context multipliers apply. Checked 14 October 2026.
Is OpenAI Decisions (GPT-6 Luna) open weight?
No. OpenAI Decisions (GPT-6 Luna) is only available through OpenAI API (public beta), OpenRouter, Vercel AI Gateway.
How fast is OpenAI Decisions (GPT-6 Luna)?
OpenAI reports About 150 ms, measured on its own hardware. JevBench gives 0.30 seconds per decision, measured end to end.
How does OpenAI Decisions (GPT-6 Luna) compare with Jev?
On JevBench's Capability Score (board v1.7.46), OpenAI Decisions (GPT-6 Luna) scores 73.5 against Jev's 77.1. Jev costs $0.042 per million input tokens; OpenAI Decisions (GPT-6 Luna) costs $0.10. JevBench is a one-person project that uses Jev as its reference, which favours cheap hosted models.