Qwen3.8-Flash-NextvsQwen3.8-27B
Across 9 shared benchmarks, Qwen3.8-Flash-Next leads overall: Qwen3.8-Flash-Next wins 9, Qwen3.8-27B wins 0, with 0 ties and an average score difference of +4.19.
Qwen3.8-Flash-Next
阿里巴巴 · 2026-08-26 · Reasoning model
Qwen3.8-27B
阿里巴巴 · 2026-08-14 · Reasoning model
Qwen3.8-Flash-Next9 wins(100%)(0%)0 winsQwen3.8-27B
Benchmark scores
Grouped by capability, sorted by largest gap within each. 9 shared benchmarks.
Coding and Software Engineer
Qwen3.8-Flash-Next 4/4| Benchmark | Qwen3.8-Flash-Next | Qwen3.8-27B | Diff |
|---|---|---|---|
| DeepSWE | 58.7016 / 30极高强度思考（工具） | 42.2025 / 30Thinking (With Tools) | +16.50 |
| NL2Repo-Bench | 48.108 / 10极高强度思考（工具） | 42.309 / 10Thinking (With Tools) | +5.80 |
| LiveCodeBench | 91.903 / 127极高强度思考（无工具） | 90.307 / 127Thinking (No Tools) | +1.60 |
| SWE-Bench Pro - Public | 62.509 / 58极高强度思考（工具） | 61.7011 / 58Thinking (With Tools) | +0.80 |
Multimodal Understanding
Qwen3.8-Flash-Next 2/2| Benchmark | Qwen3.8-Flash-Next | Qwen3.8-27B | Diff |
|---|---|---|---|
| MathVision | 95.702 / 12极高强度思考（工具） | 94.603 / 12Thinking (With Tools) | +1.10 |
| CharXiv RQ | 90.602 / 18极高强度思考（工具） | 90.203 / 18Thinking (With Tools) | +0.40 |
Agent Level Benchmark
Qwen3.8-Flash-Next 1/1| Benchmark | Qwen3.8-Flash-Next | Qwen3.8-27B | Diff |
|---|---|---|---|
| Agents' Last Exam | 24.3012 / 13极高强度思考（工具） | 20.4013 / 13Thinking (With Tools) | +3.90 |
General Evaluation
Qwen3.8-Flash-Next 1/1| Benchmark | Qwen3.8-Flash-Next | Qwen3.8-27B | Diff |
|---|---|---|---|
| GPQA Diamond | 91.7027 / 225极高强度思考（无工具） | 89.2048 / 225Thinking (No Tools) | +2.50 |
General Knowledge
Qwen3.8-Flash-Next 1/1| Benchmark | Qwen3.8-Flash-Next | Qwen3.8-27B | Diff |
|---|---|---|---|
| HLE | 35.9078 / 183极高强度思考（无工具） | 30.8090 / 183Thinking (No Tools) | +5.10 |
Specs
| Field | Qwen3.8-Flash-Next | Qwen3.8-27B |
|---|---|---|
| Publisher | 阿里巴巴 | 阿里巴巴 |
| Release date | 2026-08-26 | 2026-08-14 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | Dense |
| Parameters | 125B | 27B |
| Context length | 262144 | 256K |
| Max output | 128K | 128K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Qwen3.8-Flash-Next | Qwen3.8-27B |
|---|---|---|
| Text input | Not public | $0.5 / 1M tokens |
| Text output | Not public | $3 / 1M tokens |
| Cache read | Not public | $0.1 / 1M tokens |
One or both models have incomplete public pricing.
Summary
- Qwen3.8-Flash-Nextleads in:Coding and Software Engineer (4/4), Multimodal Understanding (2/2), Agent Level Benchmark (1/1), General Evaluation (1/1), General Knowledge (1/1)
On average across the 9 shared benchmarks, Qwen3.8-Flash-Next scores 4.19 higher.
Largest single-benchmark gap: DeepSWE — Qwen3.8-Flash-Next 58.70 vs Qwen3.8-27B 42.20 (+16.50).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.