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。

Qwen 3.6 Plus Preview
阿里巴巴
Each axis is a category average, normalized to a 100-point radar.
Relative edge: AI Agent - 工具使用 +5.3 / Relative gap: 指令跟随 -2.3
Relative edge: 指令跟随 +2.3 / Relative gap: AI Agent - 工具使用 -5.3
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 · 74.44
Best single
Qwen 3.6 Plus Preview · AIME 2026 95.30
Modality coverage
Qwen 3.6 Plus Preview · 0 modalities
Head to head
14
Benchmarks
12
Wins
2
Losses
+2.59
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.
14 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Qwen 3.6 Plus Preview | Qwen3.5-397B-A17B |
|---|---|---|
C-Eval 综合评估 | 93.30Thinking Enabled | 93.00Thinking Enabled |
GPQA Diamond 综合评估 | 90.40Thinking Enabled | 88.40Thinking Enabled |
HLE 综合评估 | 50.60Thinking Enabled | Tools | 48.30Thinking Enabled | Tools |
MMLU Pro 综合评估 | 88.50Thinking Enabled | 87.80Thinking Enabled |
LiveCodeBench 编程与软件工程 | 87.10Thinking Enabled | 83.60Thinking Enabled |
SWE-bench Multilingual 编程与软件工程 | 73.80Thinking Enabled | 69.30Thinking Enabled |
SWE-Bench Pro - Public 编程与软件工程 | 56.60Thinking Enabled | Tools | 50.90Thinking Enabled |
SWE-bench Verified 编程与软件工程 | 78.80Thinking Enabled | Tools | 76.40Thinking Enabled | Tools |
IF Bench 指令跟随 | 74.20Thinking Enabled | 76.50Thinking Enabled |
Terminal Bench 2.0 AI Agent - 工具使用 | 61.60Thinking Enabled | Tools | 52.50Thinking Enabled | Tools |
Tool Decathlon AI Agent - 工具使用 | 39.80Thinking Enabled | Tools | 38.30Thinking Enabled | Tools |
AIME 2026 数学推理 | 95.30Thinking Enabled | 91.30Thinking Enabled |
IMO-AnswerBench 数学推理 | 83.80Thinking Enabled | 80.90Thinking Enabled |
AA-LCR 长上下文能力 | 68.30Thinking Enabled | 68.70Thinking Enabled |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | Qwen 3.6 Plus Preview阿里巴巴 | Qwen3.5-397B-A17B阿里巴巴 |
|---|---|---|
Core specsRelease | 2026-03-31 | 2026-02-16 |
Context length | 1M | 256K |
Parameters | — | 397 |
Active parameters | Not provided | 17 |
Max output | 65536 | Not provided |
MoE | No | Yes |
LicenseCode Open Source | Not provided | Not provided |
Weights Open Source | Not provided | Not provided |
Commercial use | 不开源 | 免费商用授权 |
ResourcesPaper / report | Qwen3.6-Plus: Towards Real World Agents | Qwen3.5: Towards Native Multimodal Agents |

Qwen3.5-397B-A17B
阿里巴巴