See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 2 个模型的评测数据与核心参数。
Kimi K2.5
Moonshot AI
Each axis is a category average, normalized to a 100-point radar.
Relative edge: OpenClaw智能体能力综合测评 +0.8 / Relative gap: AI Agent - 信息收集 -22.6
Relative edge: AI Agent - 信息收集 +22.6 / Relative gap: OpenClaw智能体能力综合测评 -0.8
Method: for each model and benchmark, the chart first averages all scores in the current mode scope instead of taking the best score, then averages those benchmark scores within each category. Only benchmarks with at least two selected models scored are included; missing values are not counted as zero.
Best overall
Kimi K2.6 · 77.91
Best single
Kimi K2.6 · AIME 2026 96.40
Modality coverage
Kimi K2.6 · 3 modalities
Head to head
12
Benchmarks
1
Wins
11
Losses
-6.27
Average diff
Compare benchmark results across thinking modes and tool usage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Complete scores for each model/mode across selected benchmarks.
12 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Kimi K2.5 | Kimi K2.6 |
|---|---|---|
GPQA Diamond 综合评估 | 87.60Thinking Enabled | 90.50Thinking Enabled |
HLE 综合评估 | 50.20Thinking Enabled ｜ Tools | 54.00Thinking Enabled ｜ Tools |
LiveBench 综合评估 | 69.07Thinking Enabled | 72.17Thinking Enabled |
LiveCodeBench 编程与软件工程 | 85.00Thinking Enabled | 89.60Thinking Enabled |
SWE-bench Multilingual 编程与软件工程 | 73.00Thinking Enabled | 76.70Thinking Enabled ｜ Tools |
SWE-Bench Pro - Public 编程与软件工程 | 50.70Thinking Enabled ｜ Tools | 58.60Thinking Enabled ｜ Tools |
SWE-bench Verified 编程与软件工程 | 76.80Thinking Enabled ｜ Tools | 80.20Thinking Enabled ｜ Tools |
AIME 2026 数学推理 | 92.50Thinking Enabled | 96.40Thinking Enabled |
IMO-AnswerBench 数学推理 | 81.80Thinking Enabled | 86.00Thinking Enabled |
BrowseComp AI Agent - 信息收集 | 60.60Thinking Enabled ｜ Tools | 83.20Thinking Enabled ｜ Tools |
Terminal Bench 2.0 AI Agent - 工具使用 | 50.80Thinking Enabled ｜ Tools | 66.70Thinking Enabled ｜ Tools |
Claw Bench OpenClaw智能体能力综合测评 | 81.70Thinking Enabled ｜ Tools | 80.90Thinking Enabled ｜ Tools |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | Kimi K2.5Moonshot AI | Kimi K2.6Moonshot AI |
|---|---|---|
Core specsRelease | 2026-01-27 | 2026-04-20 |
Context length | 256K | 256K |
Parameters | 10000 | 10000 |
Active parameters | 320 | 320 |
Max output | 16384 | Not provided |
MoE | Yes | Yes |
Supported modes | 常规模式（Non-Thinking Mode）思考模式（Thinking Mode） | 常规模式（Non-Thinking Mode）思考模式（Thinking Mode） |
LicenseCode Open Source | Not provided | Not provided |
Weights Open Source | Not provided | Not provided |
Commercial use | 免费商用授权 | 免费商用授权 |
Modality supportText Input/Output | / | / |
Image Input/Output | / | / |
Video Input/Output | Not provided | / |
ResourcesPaper / report | Kimi K2.5: Visual Agentic Intelligence | Kimi K2.6: Advancing Open-Source Coding |
DataLearner blog | 重磅！Kimi K2.5发布，依然免费开源！原生多模态MoE架构，全球最大规模参数的开源模型之一，官方评测结果比肩诸多闭源模型！可以驱动100个子Agent执行！ | Not provided |
Kimi K2.6
Moonshot AI
