Qwen3.7-Max-PreviewvsQwen3-Max-Thinking
Across 7 shared benchmarks, Qwen3.7-Max-Preview leads overall: Qwen3.7-Max-Preview wins 7, Qwen3-Max-Thinking wins 0, with 0 ties and an average score difference of +5.39.
Qwen3.7-Max-Preview
阿里巴巴 · 2026-05-20 · Reasoning model
Qwen3-Max-Thinking
阿里巴巴 · 2026-01-26 · Reasoning model
Qwen3.7-Max-Preview7 wins(100%)(0%)0 winsQwen3-Max-Thinking
Benchmark scores
Grouped by capability, sorted by largest gap within each. 7 shared benchmarks.
General Knowledge
Qwen3.7-Max-Preview 3/3| Benchmark | Qwen3.7-Max-Preview | Qwen3-Max-Thinking | Diff |
|---|---|---|---|
| GPQA Diamond | 92.4010 / 177最高（无工具） | 87.4034 / 177thinking | +5 |
| MMLU Pro | 89.604 / 126最高（无工具） | 85.7021 / 126thinking | +3.90 |
| HLE | 53.508 / 154Thinking (With Tools) | 49.8021 / 154thinking + 使用工具 | +3.70 |
Coding and Software Engineer
Qwen3.7-Max-Preview 2/2| Benchmark | Qwen3.7-Max-Preview | Qwen3-Max-Thinking | Diff |
|---|---|---|---|
| LiveCodeBench | 91.604 / 120最高（无工具） | 85.9014 / 120thinking | +5.70 |
| SWE-bench Verified | 80.409 / 105Thinking (With Tools) | 75.3030 / 105thinking | +5.10 |
Instruction Following
Qwen3.7-Max-Preview 1/1| Benchmark | Qwen3.7-Max-Preview | Qwen3-Max-Thinking | Diff |
|---|---|---|---|
| IF Bench | 79.102 / 29最高（无工具） | 70.9011 / 29thinking + 使用工具 | +8.20 |
Math and Reasoning
Qwen3.7-Max-Preview 1/1| Benchmark | Qwen3.7-Max-Preview | Qwen3-Max-Thinking | Diff |
|---|---|---|---|
| IMO-AnswerBench | 901 / 19最高（无工具） | 83.909 / 19thinking | +6.10 |
Specs
| Field | Qwen3.7-Max-Preview | Qwen3-Max-Thinking |
|---|---|---|
| Publisher | 阿里巴巴 | 阿里巴巴 |
| Release date | 2026-05-20 | 2026-01-26 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | MoE |
| Parameters | 10000.0 | 10000.0 |
| Context length | 1M | 1000K |
| Max output | 65536 | 32768 |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Qwen3.7-Max-Preview | Qwen3-Max-Thinking |
|---|---|---|
| Text input | $2.5 / 1M tokens | 1.2 美元/100万 tokens |
| Text output | $7.5 / 1M tokens | 6 美元/100万 tokens |
| Cache read | $0.25 / 1M tokens | Not public |
| Cache write | $3.125 / 1M tokens | Not public |
Summary
- Qwen3.7-Max-Previewleads in:General Knowledge (3/3), Coding and Software Engineer (2/2), Instruction Following (1/1), Math and Reasoning (1/1)
On average across the 7 shared benchmarks, Qwen3.7-Max-Preview scores 5.39 higher.
Largest single-benchmark gap: IF Bench — Qwen3.7-Max-Preview 79.10 vs Qwen3-Max-Thinking 70.90 (+8.20).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.