跳轉到主要內容

MiniMax M1 80k

MiniMaxMiniMax開源權重MIT · 商用許可

描述

MiniMax-M1 is an open-source, large-scale reasoning model that uses a hybrid-attention architecture for efficient long-context processing. It supports up to a 1 million token context window and 80,000-token reasoning output, matching Gemini 2.5 Pro’s scale while being highly cost-effective. Its Lightning Attention mechanism reduces compute requirements to about 30% of DeepSeek R1’s, and a new reinforcement learning algorithm, CISPO, doubles convergence speed compared to other RL methods. Trained on 512 H800s over three weeks, M1 achieves near state-of-the-art results across software engineering, long-context, and tool-use benchmarks, outperforming most open models and rivaling top closed systems.

發布日期
2025-06-17
參數規模
456.0B
上下文長度
—
支援模態
text

能力雷達圖

60
general
60
coding
70
reasoning
60
science估算
60
agents
0
multimodal

缺少專門科學評測時,Science 由 LLM Stats 科學得分或推理能力估算。

排行榜排名

暫無排名資料

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Chat

TAU-bench Retail63.5%自報
Multi-Challenge44.7%自報

Factuality

SimpleQA18.5%自報

Language

MMLU-Pro81.1%自報

Long Context

OpenAI-MRCR: 2 needle 128k73.4%自報
LongBench v261.5%自報
OpenAI-MRCR: 2 needle 1M56.2%自報

Math

MATH-50096.8%自報
AIME 202486.0%自報
AIME 202576.9%自報

Reasoning

ZebraLogic86.8%自報
GPQANYU + Cohere + Anthropic (2023)70.0%自報
LiveCodeBench65.0%自報
TAU-bench Airline62.0%自報
SWE-Bench Verified56.0%自報
Humanity's Last Exam8.4%自報

AA 評測指數

(Artificial Analysis)

暫無 AA 評測資料

LLM Stats 分類評分

(LLM Stats (zeroeval))
Language
80
Legal
80
Finance
80
Healthcare
80
Math
70
Physics
70
Biology
70
Chemistry
70
Long Context
60
Reasoning
60
Structured Output
60
Frontend Development
60
General
60
Code
60
Communication
60
Tool Calling
60
Chat
50
Factuality
20
Vision
10

定價

輸入價格$0.55 / 1M tokens
輸出價格$2.2 / 1M tokens

速度

Tokens/秒52.3
首Token延遲2.34s
首回答延遲40.60s

供應商價格排行

供應商價格排行

1 個供應商

供應商輸入輸出
1MiniMax主要
$0.55
$2.2

比較該模型在不同 API 供應商之間的定價。

外部連結