MiniMax M1

MiniMax llm

Billed as the first open-weight large-scale hybrid-attention reasoning model, combining MoE with lightning attention for a native 1M-token context.

overall ⌄

35.8

98 of 125 ranked

price · $/M tokens

self-hosted

open weights, no first-party API

to run it yourself ⌄

274 GB

minimum · 1.1 TB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmarkpublishedmeasured
MMLU-Pro81.1 —
GPQA70 —
AIME 202576.9 —
SWE-bench56 —

Lineage

No recorded lineage — root or standalone model.

view in full graph →

Metadata

orgMiniMax
released2025-06-17
params456B-A45.9B
licenseapache-2.0
hf idMiniMaxAI/MiniMax-M1-80k