GPT-6 Luna
OpenAICoding model3 posts
Every GPT-6 Luna comparison on aivs.tv: 3 side-by-side posts from X against other models on the same prompt, 3 with cost or time reported. Compare coding ability, cost and speed from real tests.
Posts
3
Summary
GPT-6 Luna (OpenAI) appears in 3 posts on aivs.tv, next to DeepSeek V4.1 Flash, Claude Haiku 5.5, Grok 4.7, Claude Opus 5.5, GPT-6 Astra and GPT-6 Sol. Tasks: lava lamp: small models, cloth simulation, 4 to 4,096 faces and lava lamp, 4 models. Reported for GPT-6 Luna: 42.8 s in @bridgebench's lava lamp: small models, 13 min / 47.5K tokens in @noclipepe's cloth simulation, 4 to 4,096 faces and 43 s in @bridgebench's lava lamp, 4 models. @bridgebench picks GPT-6 Luna as the best result in lava lamp: small models.
What each post reports for GPT-6 Luna
| Post | Time | Cost | Tokens | Author’s verdict | Original |
|---|---|---|---|---|---|
| Lava lamp: small models@bridgebench | 42.8 s | — | — | Author’s pick | X post |
| Cloth simulation, 4 to 4,096 faces@noclipepe | 13 min | — | 47,500 | — | X post |
| Lava lamp, 4 models@bridgebench | 43 s | — | — | — | X post |
Only numbers stated in each post. “—” means the post doesn’t say.
Who won where
1- Lava lamp: small modelsGPT-6 Luna built the best lamp of the three: cheapest, fastest and best.— the author’s view, @bridgebench (paraphrased)
Matchups with GPT-6 Luna
6Spellings we map to GPT-6 Luna
“GPT 6 Luna”, “Luna”