AI Coding
Koi Courtyard pond, MiniMax vs DeepSeek — MiniMax M3.1 Flash vs DeepSeek V4.1 Flash 1017
Scoreboard2
- MiniMax M3.1 FlashMiniMax
- Tokens
- 88.9M
- Top; run in MiniMax Code on the author’s M Plan, 400–700 tokens/s
- DeepSeek V4.1 Flash 1017DeepSeek
- Tokens
- —
- Bottom; the author’s earlier run, archived as 2112095534123229361
Prompt
Topic · Koi pondKoi Courtyard pond, MiniMax vs DeepSeekThe post doesn’t quote the prompt wording.
The author calls the MiniMax pond a massive leap over older MiniMax models in 3D and says it ran faster than DeepSeek; no outright winner named.
@SahilPanhotra’s view (paraphrased)
MiniMax M3.1 Flash (top, in MiniMax Code) gets the same Koi Courtyard test the author gave DeepSeek V4.1 Flash 1017 (bottom) earlier: a see-through pond with rocks, swimming koi and a bamboo fountain.
The post, itemised
- Kind
- Side-by-side
- Task
- Koi Courtyard pond, MiniMax vs DeepSeek
- Topic
- Koi pond
- Category
- 3D & Shaders
- Mode
- Agentic
- Posted
- Video length
- 34 s
- Likes
- 24
- Views
- 611
- Bookmarks
- 2
- Sponsored
- No
NotesThe two runs were made at different times and in different harnesses. The post also repeats MiniMax’s current plan offers.
Cost & time, as stated in the postAs stated: MiniMax M3.1 Flash used 88.94M tokens (86.16M cache reads), $0 extra on the M Plan; speed 400–700 tokens/s vs about 250–300 for DeepSeek.
Related
Posted by @SahilPanhotra on X. aivs.tv claims no ownership: it hosts a copy for fast playback and links to the post on X. Request removal
