MiniMax-M2.7-NVFP4

Home/Models/nvidia/MiniMax-M2.7-NVFP4

nvidia/MiniMax-M2.7-NVFP4

kind: nvfp4
quant: nvfp4 (4-bit)
family: minimax
modality: text-generation
root model: MiniMaxAI/MiniMax-M2.7
provider: nvidia
license: other
language: —
context: —
library: Model Optimizer
vRAM (GB): —
size on disk: —
downloads: 73,704
likes: 66
created: —
last modified: 2026-04-24T21:40:54+00:00 (2026)
huggingface: https://huggingface.co/nvidia/MiniMax-M2.7-NVFP4
tags: Model Optimizersafetensorsminimax_m2nvidiaModelOptMiniMaxquantizedNVFP4nvfp4text-generationconversationalcustom_codebase_model:MiniMaxAI/MiniMax-M2.7base_model:quantized:MiniMaxAI/MiniMax-M2.7license:other8-bitmodeloptdeploy:azureregion:us

benchmark	category	score	conf	evidence
IFEval	instruction-following	90.4	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
AIME 2025	math	88.8	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
GPQA Diamond	reasoning	85.7	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
MMLU-Pro	knowledge	81.7	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
IFBench	instruction-following	72.8	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
LiveCodeBench	coding	58.2	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`
SciCode	coding	48.7	high	`\| NVFP4 \| 0.904 \| 0.817 \| 0.857 \| 0.582 \| 0.487 \| 0.888 \| 0.728 \| 0.728 \|`

none scraped.