HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

MiniMax-M2.5

MiniMaxAI/MiniMax-M2.5

kind
fp8
quant
fp8 (8-bit)
family
minimax
modality
text-generation
root model
MiniMaxAI/MiniMax-M2.5
provider
MiniMaxAI
license
other
language
—
context
—
library
transformers
vRAM (GB)
—
size on disk
—
downloads
728,181
likes
1,500
created
—
last modified
2026-03-10T13:42:23+00:00 (2026)
huggingface
https://huggingface.co/MiniMaxAI/MiniMax-M2.5
tags
transformerssafetensorsminimax_m2text-generationconversationalcustom_codelicense:othereval-resultsendpoints_compatiblefp8deploy:azureregion:usdeploy:sagemaker

Benchmark scores (5)

benchmarkcategoryscorecontextconfevidence
AIME 2025math86.3AIME25medium| AIME25 | 86.3 | 83.0 | 88.0 | 91.0 | 95.6 | 96.0 | 98.0 |
GPQAreasoning85.2GPQA-Dmedium| GPQA-D | 85.2 | 83.0 | 83.0 | 87.0 | 90.0 | 91.0 | 90.0 |
IFBenchinstruction-following70.0IFBenchmedium| IFBench | 70.0 | 70.0 | 57.0 | 58.0 | 53.0 | 70.0 | 75.0 |
SciCodecoding44.4SciCodemedium| SciCode | 44.4 | 41.0 | 45.0 | 50.0 | 52.0 | 56.0 | 52.0 |
Humanity's Last Examknowledge19.4HLE w/o toolsmedium| HLE w/o tools | 19.4 | 22.2 | 17.3 | 28.4 | 30.7 | 37.2 | 31.4 |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.