| Model profile | ||
|---|---|---|
| Model | Size | Approximate size |
| MiniMax M3.1 FlashModel source: provider1,000,000 tokens max native context | 428Bactive 23B | 256.8 GB4-bit quantization |
| Compatible hardware Estimated speeds, not benchmark results: calculated from memory bandwidth and model size. Real results can differ significantly because there is no precise formula for deriving LLM generation speed from hardware specifications alone. | ||||
|---|---|---|---|---|
| Hardware | Memory | Bandwidth | vLLM | Estimated generation |
| DGX H200 | 1128 GBGPU HBM3e | 4800 GB/s | yes | 119 tok/s4-bit quantization |
| DGX Station | 748 GBCoherent Memory | 7100 GB/s | yes | 163 tok/s4-bit quantization |
| ET900N G3 | 748 GBCoherent Memory | 7100 GB/s | yes | 163 tok/s4-bit quantization |
| DGX B200 | 1440 GBGPU HBM3e | 8000 GB/s | yes | 178 tok/s4-bit quantization |
| GB200 NVL72 | 13400 GBGPU HBM3e | 8000 GB/s | yes | 178 tok/s4-bit quantization |
| GB300 NVL72 | 20000 GBGPU HBM3e | 8000 GB/s | yes | 178 tok/s4-bit quantization |
MiniMax M3.1 Flash and Claude Opus 5.5 were given three tasks from a 20,000-line Go project, passing every hidden check with costs of approximately $0.65 and $7 respectively.
Compares MiniMax M3.1 Flash and GPT-6.1 Sol generating a playable Call of Duty-style 3D FPS game using Three.js from scratch with identical prompts.
Tests MiniMax M3.1-Flash-Preview via mcode on Ubuntu using drone vision decisions, a Blender and Godot game, and debugging a legacy DB2 materialized query table.
Tests MiniMax M3.1 Flash Preview via MiniMax Code on AI coding, games, 3D environments, and websites, comparing speed and output against GPT-6 Sol and Fable 5.1.