Across 11 shared benchmarks, Qwen 3.6 Plus Preview leads overall: Qwen 3.6 Plus Preview wins 11, Kimi K2.5 wins 0, with 0 ties and an average score difference of +3.90.
阿里巴巴 · 2026-03-31 · AI model
Moonshot AI · 2026-01-27 · Multimodal model
Grouped by capability, sorted by largest gap within each. 11 shared benchmarks.
| Benchmark | Qwen 3.6 Plus Preview | Kimi K2.5 | Diff |
|---|---|---|---|
| SWE-Bench Pro - Public | 56.609 / 36Thinking (With Tools) | 50.7025 / 36Thinking (With Tools) | +5.90 |
| LiveCodeBench | 87.109 / 118Thinking (No Tools) | 8514 / 118Thinking (No Tools) | +2.10 |
| SWE-bench Verified | 78.8016 / 103Thinking (With Tools) | 76.8022 / 103Thinking (With Tools) | +2 |
| SWE-bench Multilingual | 73.805 / 17Thinking (No Tools) | 738 / 17Thinking (No Tools) | +0.80 |
| Benchmark | Qwen 3.6 Plus Preview | Kimi K2.5 | Diff |
|---|---|---|---|
| MMLU Pro | 88.504 / 124Thinking (No Tools) | 78.5064 / 124Thinking (No Tools) | +10 |
| GPQA Diamond | 90.4015 / 175Thinking (No Tools) | 87.6031 / 175Thinking (No Tools) | +2.80 |
| HLE | 50.6014 / 149Thinking (With Tools) | 50.2017 / 149Thinking (With Tools) | +0.40 |
| Benchmark | Qwen 3.6 Plus Preview | Kimi K2.5 | Diff |
|---|---|---|---|
| AIME 2026 | 95.302 / 14Thinking (No Tools) | 92.5010 / 14Thinking (No Tools) | +2.80 |
| IMO-AnswerBench | 83.809 / 17Thinking (No Tools) | 81.8012 / 17Thinking (No Tools) | +2 |
| Benchmark | Qwen 3.6 Plus Preview | Kimi K2.5 | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 61.6014 / 43Thinking (With Tools) | 50.8030 / 43Thinking (With Tools) | +10.80 |
| Benchmark | Qwen 3.6 Plus Preview | Kimi K2.5 | Diff |
|---|---|---|---|
| AA-LCR | 68.306 / 13Thinking (No Tools) | 6510 / 13Thinking (No Tools) | +3.30 |
| Field | Qwen 3.6 Plus Preview | Kimi K2.5 |
|---|---|---|
| Publisher | 阿里巴巴 | Moonshot AI |
| Release date | 2026-03-31 | 2026-01-27 |
| Model type | AI model | Multimodal model |
| Architecture | Dense | MoE |
| Parameters | 0.0 | 10000.0 |
| Context length | 1M | 256K |
| Max output | 65536 | 16384 |
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Qwen 3.6 Plus Preview | Kimi K2.5 |
|---|---|---|
| Text input | $0.5 / 1M tokens | 0.6 美元/100 万tokens |
| Text output | $3 / 1M tokens | 3 美元/100 万tokens |
| Cache read | $0.05 / 1M tokens | 0.1 美元/100 万tokens |
| Cache write | $0.625 / 1M tokens | Not public |
On average across the 11 shared benchmarks, Qwen 3.6 Plus Preview scores 3.90 higher.
Largest single-benchmark gap: Terminal Bench 2.0 — Qwen 3.6 Plus Preview 61.60 vs Kimi K2.5 50.80 (+10.80).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.
