Qwen3.5-27BvsQwen3-32B
Across 4 shared benchmarks, Qwen3.5-27B leads overall: Qwen3.5-27B wins 4, Qwen3-32B wins 0, with 0 ties and an average score difference of +158.37.
Qwen3.5-27B
阿里巴巴 · 2026-02-25 · Reasoning model
Qwen3-32B
阿里巴巴 · 2025-04-28 · Reasoning model
Qwen3.5-27B4 wins(100%)(0%)0 winsQwen3-32B
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
Coding and Software Engineer
Qwen3.5-27B 2/2| Benchmark | Qwen3.5-27B | Qwen3-32B | Diff |
|---|---|---|---|
| CodeForces | 1,89915 / 16Thinking (No Tools) | 1,35316 / 16Normal (No Tools) | +546 |
| LiveCodeBench | 80.7027 / 120Thinking (With Tools) | 31.30114 / 120Normal (No Tools) | +49.40 |
General Knowledge
Qwen3.5-27B 2/2| Benchmark | Qwen3.5-27B | Qwen3-32B | Diff |
|---|---|---|---|
| GPQA Diamond | 85.5047 / 178Thinking (No Tools) | 54.60148 / 178Normal (No Tools) | +30.90 |
| C-Eval | 90.506 / 9Thinking (No Tools) | 83.309 / 9Normal (No Tools) | +7.20 |
Specs
| Field | Qwen3.5-27B | Qwen3-32B |
|---|---|---|
| Publisher | 阿里巴巴 | 阿里巴巴 |
| Release date | 2026-02-25 | 2025-04-28 |
| Model type | Reasoning model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | 27B | 32B |
| Context length | 1010K | 128K |
| Max output | 248320 | 16K |
Summary
- Qwen3.5-27Bleads in:Coding and Software Engineer (2/2), General Knowledge (2/2)
On average across the 4 shared benchmarks, Qwen3.5-27B scores 158.37 higher.
Largest single-benchmark gap: CodeForces — Qwen3.5-27B 1,899 vs Qwen3-32B 1,353 (+546).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.