Moonshot AI
Moonshot's K2 base model with exceptional coding and agent capabilities, built on a MoE architecture with strong instruction following.
54.1
Quality Score
1237
Arena ELO
Undisclosed
Parameters
256K
Context
Use this section to answer one simple question first: how much outside evidence do we have that this model performs well? Structured benchmark scores appear first, then official provider evidence, then live arena signal.
This model has normalized benchmark rows, so scores here are directly comparable across benchmark sources.
Sign in to join the discussion
0
Downloads
0
Likes
Sep 2025
Released
These are recent benchmark or leaderboard claims from official provider sources. They are useful for freshness and context, but they are not treated the same as normalized independent benchmark rows.
Kimi K2 Benchmark Update
Quality: 19.7/100 | Price: $1.002/M tokens | Output: 0 tok/s | MMLU: 0.824% | HumanEval: 0.556%
View sourcekimi-k2-instruct - SWE-Bench Verified
SWE-Bench Verified resolved rate 53.4
View sourceKimi K2 Benchmark Update
Quality: 19.7/100 | Price: $1.002/M tokens | Output: 0 tok/s | MMLU: 0.824% | HumanEval: 0.556%
View sourceKimi K2 Benchmark Update
Quality: 19.7/100 | Price: $1.002/M tokens | Output: 0 tok/s | MMLU: 0.824% | HumanEval: 0.556%
View sourceKimi K2 Benchmark Update
Quality: 19.7/100 | Price: $1.002/M tokens | Output: 0 tok/s | MMLU: 0.824% | HumanEval: 0.556%
View source模型列表 - Kimi API 开放平台
Navigation 开始使用 模型列表 使用指南 API 接口说明 产品定价 常见问题 资源 开始使用 快速开始 模型列表 Kimi K3 模型 Kimi K2.7 Code 模型 Kimi K2.6 模型 模型能力 思考模型 推理强度 多轮对话 流式输出 JSON Mode Partial Mode 视觉输入 上下文缓存 动态加载工具 工具调用 基础介绍 联网搜索 官方工具列表 工具调用约束 ModelScope MCP 工具调用最佳实践 核心工作流 response_format 自动断线重连 文件问答指南 Batch API 指南 生态集成 Kimi Code CLI OpenClaw Claude Code OpenCode Hermes Agent Codex 调试与运维 调试工具 开发工作台调试 组织管理最佳实践 在此页面 多模态模型 生成模型 Moonshot V1 已下线模型 开始使用 模型列表 复制页面 复制页面 查看 Kimi 开放平台当前可用的多模态、编程与 Moonshot V1 模型，以及已下线模型的迁移提示。 复制页
View source1237
ELO Score
1226 - 1248
95% Confidence
+/-11 points
3.8K
Battles
Aug 19, 2026
Last Updated