GPT-6 Luna vs GPT-6.1 Terra
aka Luna vs OpenAI GPT-6.1 Terra
- CheaperGPT-6 Luna3/3 postsmedian gap ~2.7×
- FasterGPT-6.1 Terra2/2 postsmedian gap ~2.8×
As stated in each post, not a controlled benchmark.
Posts
5Chrome dino game, four GPT models: GPT-6.1 Terra, GPT-6 Luna, GPT-6.1 Sol, GPT-6 Astra. Posted by @noclipepe.
Basic scene test, GPT-6 Luna vs GPT-6.1 Terra: GPT-6 Luna, GPT-6.1 Terra. Posted by @VulKan42069.
Pelican with a dog on its back, GPT lineup SVG: GPT-6.1 Terra, GPT-6 Luna, GPT-6.1 Sol. Posted by @trypalomino.
3D office from a floor plan, four GPT models: GPT-6.1 Terra, GPT-6.1 Sol, GPT-6 Astra, GPT-6 Luna. Posted by @koharu_ai_diary.
Same prompt, separate runs
3Not a controlled comparison: prompts, tools and time differ.
About this matchup
GPT-6 Luna vs GPT-6.1 Terra in real side-by-side tests: 5 posts from X where both got the same prompt; 4 posts report cost or time. Judge coding ability from the actual outputs.
GPT-6 Luna vs GPT-6.1 Terra: which is better for coding?
Which is better for coding, GPT-6 Luna or GPT-6.1 Terra?
None of the 5 posts here names one of the two as the better result, so judge the outputs above for the work you care about.
Which is cheaper, GPT-6 Luna or GPT-6.1 Terra?
3 posts state a cost for both: GPT-6 Luna was cheaper in all 3 (median gap ~2.7×). These are the costs the posts state for each run, not list prices.
Which is faster, GPT-6 Luna or GPT-6.1 Terra?
2 posts state a time for both: GPT-6.1 Terra was faster in both (median gap ~2.8×). Times are as stated in each post; tools and settings differ between posts.
Summary
GPT-6 Luna and GPT-6.1 Terra appear together in 5 posts on X: 3D take on an app character, four GPT models (@floriv__), chrome dino game, four GPT models (@noclipepe), basic scene test, GPT-6 Luna vs GPT-6.1 Terra (@VulKan42069), pelican with a dog on its back, GPT lineup SVG (@trypalomino) and 3D office from a floor plan, four GPT models (@koharu_ai_diary). 4 of them are part of larger lineups with other models. In @noclipepe's chrome dino game, four GPT models, GPT-6 Luna was cheaper ($0.09 vs $0.17) and GPT-6.1 Terra was faster (4.3 min vs 10.2 min), as reported in the post. In @VulKan42069's basic scene test, GPT-6 Luna vs GPT-6.1 Terra, GPT-6 Luna was cheaper ($0.056 vs $0.15) and GPT-6.1 Terra was faster (5.7 min vs 18.3 min), as reported in the post. In @trypalomino's pelican with a dog on its back, GPT lineup SVG, GPT-6 Luna was cheaper ($0.0075 vs $0.05) and GPT-6.1 Terra used fewer tokens (7.6K tokens vs 11.9K tokens), as reported in the post.
What each post reports
| Post | Model | Time | Cost | Tokens | Post |
|---|---|---|---|---|---|
| 3D take on an app character, four GPT models@floriv__ | GPT-6.1 Terra | — | — | — | X post |
| 3D take on an app character, four GPT models@floriv__ | GPT-6 Luna | — | — | — | X post |
| Chrome dino game, four GPT models@noclipepe | GPT-6.1 Terra | 4.3 min | $0.17 | — | X post |
| Chrome dino game, four GPT models@noclipepe | GPT-6 Luna | 10.2 min | $0.09 | — | X post |
| Basic scene test, GPT-6 Luna vs GPT-6.1 Terra@VulKan42069 | GPT-6 Luna | 18.3 min | $0.056 | — | X post |
| Basic scene test, GPT-6 Luna vs GPT-6.1 Terra@VulKan42069 | GPT-6.1 Terra | 5.7 min | $0.15 | — | X post |
| Pelican with a dog on its back, GPT lineup SVG@trypalomino | GPT-6.1 Terra | — | $0.05 | 7,600 | X post |
| Pelican with a dog on its back, GPT lineup SVG@trypalomino | GPT-6 Luna | — | $0.0075 | 11,900 | X post |
| 3D office from a floor plan, four GPT models@koharu_ai_diary | GPT-6.1 Terra | 48 s | — | — | X post |
| 3D office from a floor plan, four GPT models@koharu_ai_diary | GPT-6 Luna | — | — | — | X post |
Only numbers stated in each post. “—” means the post doesn’t say.
Quick facts
| Model | Vendor | Released | Input $/1M tokens | Output $/1M tokens | Context (tokens) | Official links |
|---|---|---|---|---|---|---|
| GPT-6 Luna | OpenAI | |||||
| GPT-6.1 Terra | OpenAI | 2026-10-26 | Announcement |
From the vendors’ official pages, checked 2026-10-26. Standard API list prices; each value links its source and a blank cell could not be verified. The archive’s own numbers above are what each post states.










