MiniMax M3.1 Flash scores 74% on DeepSWE and 91% on BabyVision
The poster confirms Space Bunny was an early M3.1 Flash build, cites official scores of 74% DeepSWE, 91% BabyVision and 62% Automation-Bench, and says hands-on results were solid at one-third M3's cost.
What a comeback by @MiniMax_AI! It's been 140 days since MiniMax M3! Turns out, Space Bunny was an earlier version of MiniMax-M3.1-Flash, and the official release is even better! (More results coming soon !) New M3.1-Flash scores 74% on DeepSWE for coding, 91% on BabyVision for visual reasoning, and 62% on Automation-Bench for automation tasks. Best part? It delivers frontier-level intelligence at 1/3 the cost of M3 and 3x the speed! My own testing shows it's surprisingly good, delivering really solid results. It's a big jump over M3 and a strong contender among open models. MiniMax has my attention again.
This post quotes another post: x.com/minimax_ai/status/2112150867868541221