MiniMax M3 has landed in the Arena and has moved the Pareto frontier!
TL;DR · AI Summary
MiniMax M3 has debuted in Code Arena, ranking #7 in the frontend track with a score of 1,531, tying with GLM-5.1. It advances the Pareto frontier in its price class at $0.60/ $2.40 per Mtoken.
Key Takeaways
- Code Arena frontend #7, score 1,531, tied with GLM-5.1.
- Advances Pareto frontier at $0.60/ input, $2.40/ output per Mtoken.
- First open-weights model combining coding, agentic, and sparse attention, with 1
Outline
Jump quickly between sections.
MiniMax M3 debuts in Code Arena and moves the Pareto frontier in its price class.
Code Arena frontend #7 with score 1,531, tied with GLM-5.1.
Advances the Pareto frontier at $0.60/ input, $2.40/ output per Mtoken.
First open-weights model combining coding, agentic, and sparse attention for 1M token context.
SWE-Bench Pro 59.0%, Terminal Bench 2.1 66.0%, SWE-fficiency 34.8%.
Utilizes sparse attention to scale context and improve efficiency.
Mindmap
See how the topics connect at a glance.
查看大纲文本（无障碍 / 无 JS 友好）
- MiniMax M3与Code Arena突破
- 排名与基准
- 前端第7，得分1531，与GLM-5.1并列
- 价格与前沿
- 每Mtoken $0.60/输入、$2.40/输出
- 技术亮点
- 开源权重
- 融合编码与代理
- 稀疏注意力，上下文至1M token
- 基准成绩
- SWE-Bench Pro 59.0%
- Terminal Bench 2.1 66.0%
- SWE-fficiency 34.8%
Highlights
Key sentences worth saving and sharing.
Code Arena frontend #7 with score 1,531, tied with GLM-5.1 (tied ranking + specific score).
Advances the Pareto frontier at $0.60/ input, $2.40/ output per Mtoken (price comparison + frontier突破).
First open-weights model combining coding, agentic, and sparse attention for 1M token context (technical integration + quantitative metric).
Their latest model ranks #7 for Code Arena: Frontend, scoring 1531, it is neck and neck with GLM-5.1. It moves the Pareto frontier in its price class at $0.60 input/$2.40 output per Mtoken.
Congrats to the https://t.co/GbFqAWVXj0" / X
MiniMax M3 has landed in the Arena and has moved the Pareto frontier! Their latest model ranks #7 for Code Arena: Frontend, scoring 1531, it is neck and neck with GLM-5.1. It moves the Pareto frontier in its price class at $0.60 input/$2.40 output per Mtoken. Congrats to the
team on this achievement!
Quote
MiniMax (official)
@MiniMax_AI
Jun 1
Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities - Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas - MiniMax Sparse Attention scales context to 1M -
