MiniMax Research · 2026-02-12精选模型评分81
MiniMax 发布 M2.5 模型，SWE-Bench Verified 得分 80.2%
MiniMax M2.5: Built for Real-World Productivity.2026-02-12Learn More >
导读
MiniMax 发布新模型 MiniMax-M2.5，在 SWE-Bench Verified 上得分 80.2%，Multi-SWE-Bench 得分 51.3%，BrowseComp 得分 76.3%。
推荐理由
官方给出 M2.5 在编码、搜索与办公任务上的基准分数和每任务成本，读者可据此比较它与 Claude、GPT 系列的实际差距。
分类置信度
高
- 官方或准官方来源，且无冲突类别。
- 正文、可信日期、主分类和证据摘录完整。
来源级别：官方一手
事件类型：研究发布
证据摘录
- “Today we're introducing our latest model, MiniMax-M2.5.”
- “We’re releasing two versions of the model, M2.5 and M2.5-Lightning, that are identical in capability but differ in speed.”
- “M2.5-Lightning has a steady throughput of 100 tokens per second, which is two times faster than other frontier models, and costs $0.3 per million input tokens and $2.4 per million output tokens.”
- “M2.5, which has a throughput of 50 tokens per second, costs half that.”
- “Both model versions support caching.”
来源：MiniMax Research · minimax.io