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: none clear / Relative gap: AI Agent - 工具使用 -10.8
Relative edge: AI Agent - 工具使用 +10.8 / Relative gap: none clear
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
Qwen 3.6 Plus Preview · 74.43
Best single
Qwen 3.6 Plus Preview · AIME 2026 95.30
Modality coverage
Qwen 3.6 Plus Preview · 3 modalities
Head to head
13
Benchmarks
0
Wins
13
Losses
-3.51
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.
13 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Kimi K2.5 | Qwen 3.6 Plus Preview |
|---|---|---|
GPQA Diamond 综合评估 | 87.60Thinking Enabled | 90.40Thinking Enabled |
HLE 综合评估 | 50.20Thinking Enabled ｜ Tools | 50.60Thinking Enabled ｜ Tools |
LiveBench 综合评估 | 69.07Thinking Enabled | 70.85Standard Mode |
MMLU Pro 综合评估 | 78.50Thinking Enabled | 88.50Thinking Enabled |
LiveCodeBench 编程与软件工程 | 85.00Thinking Enabled | 87.10Thinking Enabled |
SWE-bench Multilingual 编程与软件工程 | 73.00Thinking Enabled | 73.80Thinking Enabled |
SWE-Bench Pro - Public 编程与软件工程 | 50.70Thinking Enabled ｜ Tools | 56.60Thinking Enabled ｜ Tools |
SWE-bench Verified 编程与软件工程 | 76.80Thinking Enabled ｜ Tools | 78.80Thinking Enabled ｜ Tools |
AIME 2026 数学推理 | 92.50Thinking Enabled | 95.30Thinking Enabled |
IMO-AnswerBench 数学推理 | 81.80Thinking Enabled | 83.80Thinking Enabled |
Terminal Bench 2.0 AI Agent - 工具使用 | 50.80Thinking Enabled ｜ Tools | 61.60Thinking Enabled ｜ Tools |
AA-LCR 长上下文能力 | 65.00Thinking Enabled | 68.30Thinking Enabled |
LongBench v2 长上下文能力 | 61.00Standard Mode | 62.00Standard Mode |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | Kimi K2.5Moonshot AI | Qwen 3.6 Plus Preview阿里巴巴 |
|---|---|---|
Core specsRelease | 2026-01-27 | 2026-03-31 |
Context length | 256K | 1M |
Parameters | 10000 | — |
Active parameters | 320 | Not provided |
Max output | 16384 | 65536 |
MoE | Yes | No |
Supported modes | 常规模式（Non-Thinking Mode）思考模式（Thinking Mode） | No mode data |
LicenseCode Open Source | Not provided | Not provided |
Weights Open Source | Not provided | Not provided |
Commercial use | 免费商用授权 | 不开源 |
Modality supportText Input/Output | / | / |
Image Input/Output | / | / |
Video Input/Output | Not provided | / |
ResourcesPaper / report | Kimi K2.5: Visual Agentic Intelligence | Qwen3.6-Plus: Towards Real World Agents |
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 |
Qwen 3.6 Plus Preview
阿里巴巴
