MiniMax-M3
开始使用
API 接入
Tempr Gateway API文档
目录部署标识:minimax/MiniMax-M3。具体区域、权限与价格以服务商为准。
Eden AI API文档
目录部署标识:minimax/MiniMax-M3。具体区域、权限与价格以服务商为准。
MiniMax (minimax.io) API文档
目录部署标识:MiniMax-M3。具体区域、权限与价格以服务商为准。
公开成绩与计算依据
综合参考分根据已公开的评测资料整理。查看发布日期依据
综合参考分 45.1 · 17项基准
推理求解 59.0
GPQA Diamond · 0.929292929292929 比例
Artificial Analysis · 独立评测 · Diamond
MiniMax-M3 · Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond
Terminal-Bench-Science · 0.00476190476190476 比例
Artificial Analysis · 独立评测 · 0.1.0
MiniMax-M3 · Artificial Analysis current published protocol; terminal-bench-science
HLE · 0.16 比例
Artificial Analysis · 独立评测 · Diamond
MiniMax-M3 · Artificial Analysis current published protocol; hle-diamond
CritPt · 0.28 比例
Artificial Analysis · 独立评测 · 2 Hard
MiniMax-M3 · Artificial Analysis current published protocol; critpt-v2-hard
工具与任务执行 28.0
tau3-Banking · 0.152577319587629 比例
Artificial Analysis · 独立评测 · 1.0.1
MiniMax-M3 · Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking
GDPval-AA · 1249.95 分
Artificial Analysis · 独立评测 · 2.1
MiniMax-M3 · Artificial Analysis current published protocol; gdpval-aa-normalized
Terminal-Bench · 1.01%
Vals AI · 独立评测 · 4.0
MiniMax-M3 / Mini-SWE-agent · Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.
AutomationBench · 0.21251257600749407 比例
Artificial Analysis · 独立评测 · current published task suite
MiniMax-M3 · Artificial Analysis current published protocol; automationbench-aa
AA-Briefcase · 1089.09 分
Artificial Analysis · 独立评测 · 1.1
MiniMax-M3 · Artificial Analysis current published protocol; aa-briefcase
EnterpriseOps-Gym · 0.3210981796478663 比例
Artificial Analysis · 独立评测 · current published task suite
MiniMax-M3 · Artificial Analysis current published protocol; enterprise-ops-gym-aa
AA-AnalystAgent · 0.1 比例
Artificial Analysis · 独立评测 · current AA pass^5
MiniMax-M3 · Artificial Analysis current published protocol; aa-analyst-agent
编程 34.5
FrontierCode · 0.219031313131313 比例
Artificial Analysis · 独立评测 · 2.0
MiniMax-M3 · Artificial Analysis current published protocol; frontiercode-2-0
Vals Code Migration · 20.90%
Vals AI · 独立评测 · 1.1
MiniMax-M3 / Published Code Migration 1.1 sandbox · Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.
语言与信息处理 60.1
IFBench · 0.828571428571429 比例
Artificial Analysis · 独立评测 · current published task suite
MiniMax-M3 · Artificial Analysis current published protocol; ifbench
GDP.pdf · 0.098 比例
Artificial Analysis · 独立评测 · current published task suite
MiniMax-M3 · Artificial Analysis current published protocol; gdp-pdf
AA-Omniscience · 0.44866666666666666 比例
Artificial Analysis · 独立评测 · 1.1
MiniMax-M3 · Artificial Analysis current published protocol; omniscience-accuracy
AA-Omniscience · 0.08645707376058043 比例
Artificial Analysis · 独立评测 · 1.1
MiniMax-M3 · Artificial Analysis current published protocol; omniscience-non-hallucination
多模态理解 44.1
MMMU-Pro · 0.785549132947977 比例
Artificial Analysis · 独立评测 · Standard 10
MiniMax-M3 · Artificial Analysis current published protocol; mmmu-pro