MiniMax-M1 — benchmark results

MiniMax's open-weight 456B MoE reasoning model (45.9B active) using hybrid lightning attention to support a 1M-token context (June 2025). Provider: MiniMax. Released 2025-06-01. Access: Open.

Unified ELO 1497 ± 45, rank #810 of 1776 rated models, from 15 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Context Arena MRCR (2-needle)30.4AUC@1M (%)77.5
SnakeBench24.1TrueSkill Rating60.6
Wolfram LLM Benchmarking Project43.7Correct Functionality (%)57.2
BenchTable52.3Total Score (%)56.7
Chatbot Arena (Text)1364Elo55.6
LLM2014 Logic 2025-0643.83Median Score52.8
AI Chess Leaderboard (Reasoning)653Elo48.7
NYT Connections Older Models21.3Score (%)34.3
LLM2014 Logic 2025-0928.28Median Score28.9
LLM2014 Logic 2025-1028.28Median Score28.6
AI Chess Leaderboard (Continuation)456Elo26.7
LLM2014 Logic 2025-0830.58Median Score22.7

Interactive version: theaggregate.ai/model?slug=minimax-m1 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.