WHAT NEW IN X10.28 WED

AI collections from X · 22 collections · 58,924 posts

minimax-m3.1VIDEO

Pelican Mafia test: MiniMax M3.1 Flash max takes 2h24m and 76.3M tokens vs DeepSeek V4.1 Flash's 41 minutes

On the same test, M3.1 Flash Max in OpenCode produced 3341 lines using 76.3M tokens over 2h24m, while DeepSeek V4.1 Flash 1017 Max produced 5369 lines with 33.6M tokens in 41m10s; the author recommends just picking DeepSeek.

VulKan
@VulKan42069Archived 7 days agoOriginally posted 2026.10.21
VulKan @VulKan42069Originally posted 2026.10.21
Open original on X

[Pelican Mafia Test 🕵️‍♂️] MiniMax M3.1 Flash Max - OpenCode: 3341 lines, 76.3M tokens, 2h 24m DeepSeek V4.1 Flash 1017 Max - DS API: 5369 lines, 33.6M tokens, 41m 10s If you're still puzzled to choose, just go for deepseek V4.1 Flash 👍

x.com/vulkan42069/status/2112584086640632280Archived 2026.10.21 12:04