Alibaba
Qwen3.8 27B (medium)
Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability.
Qwen model releasesModel Snapshot
TypeLanguage
ReleaseAug 14, 2026
Context1,000,000
Tagsreasoning, tool-use, vision, implicit-caching
Overall
Rank #134 across 687
27.6
Coding
Rank #125 across 330
56.1
Blended Price
Rank #231 across 432
$1.13/M
Strength Profile
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Benchmark Percentiles
Higher bars mean stronger relative placement.
Model Details
Runtime
Output Speed47.2 tok/s
First Token1.20s
Max Output131,072
Pricing
Blended$1.13/M
Input$0.500/M
All Benchmarks
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 27.6 | #134 |
| Artificial Analysis Coding Index | coding | 56.1 | #125 |
| GPQA | reasoning | 84.5% | #148 |
| Humanity's Last Exam | reasoning | 14.1% | #242 |
| SciCode |