MiniMax-M3-AWQ-1M-NVFP4-KV-4x-DGX-Spark. MiniMax-M3-AWQ serving 1,048,576-token (1M) context on 4x DGX Spark (GB10) via a 4-bit nvfp4 KV cache. Built on CosmicRaisins' M3 recipe.

github.com/tonyd2wild/MiniMax-M3-AWQ-1M-NVFP4-KV-4x-DGX-Spark

Vaya's read on this project

Problem, audience, market, and the verdict — sign in to see it.

Updates

No recent activity.