Unsloth's 2-bit Quantization Runs MiniMax M3.1 Flash Locally
Unsloth released UD3 GGUF quantization for MiniMax's 428B MoE model M3.1-Flash, with the 2-bit version retaining 91.5% accuracy while running locally in about 144GB, and MiniMax officially shared the results.
2026-10-26 ~ 2026-10-27 · 2 related posts
- Unsloth Ships MiniMax-M3.1-Flash Quants: 2-bit Retains 91.5% Accuracy, Runs in 150GB RAM — danielhanchen · 2026-10-26
- Unsloth squeezes 428B MiniMax M3.1 Flash into 143.9GB 2-bit GGUF with 91.5% accuracy — danielhanchen · 2026-10-27