Alibaba
Qwen3.8 27B
Dense 27B vision-language model for coding, agent tasks, and image and video understanding
alibaba/qwen3.8-27b
Context
262.1K
Output
32.8K
Released
2026-08-14
Knowledge cutoff
—
Types
Reasoning
Tool use
Structured output
Attachments
Open weights
Provider pricing
- Price
- $0.50 / $3.00
- Context
- 1M
- Status
- Stable
- Price
- $0.62 / $4.40
- Context
- 262.1K
- Status
- Stable
Benchmarks
SWE-bench Pro
61.7 · resolved
Terminal-Bench
73 · score
NL2Repo
42.3 · score
DeepSWE
42.2 · score
QwenSWEBench
79 · avg@3
CoWorkBench
70.7 · score
JobBench
33.4 · score
Agents' Last Exam
20.4 · pass@1
Agents' Last Exam
42.9 · score
IFBench
79.5 · score
GPQA Diamond
89.2 · score
Humanity's Last Exam
30.8 · score
LiveCodeBench
90.3 · score
OSWorld-Verified
84.3 · score
WebArena-Verified
64.8 · score
AndroidWorld
81.9 · score
SWE-MM
38.6 · score
RealWorldQA
85.9 · score
ERQA
65.5 · score
RecreationBench
47.1 · score
ClawEval-MM
57.4 · pass@3
ClawEval-MM
56.9 · average score
Vision2Web
62.9 · score
MathVision
90 · score
MathVision
94.6 · score
BabyVision
65.7 · score
BabyVision
85.6 · score
CharXiv Reasoning
83.7 · score
CharXiv Reasoning
90.2 · score
OmniDocBench
91.1 · score
From the same lab
Qwen 3.8 Max Prime
alibaba/qwen3.8-max-prime
Qwen3.8 Omni Flash
alibaba/qwen3.8-omni-flash
Qwen-Image-2.1
alibaba/qwen-image-2.1
Qwen3.8 Max 0902
alibaba/qwen3.8-max-0902
Qwen3.8 Flash Next
alibaba/qwen3.8-flash-next
Qwen3.8 Flash
alibaba/qwen3.8-flash
Qwen3.8 2.4T A95B
alibaba/qwen3.8-2.4t-a95b
Qwen3.8 Max
alibaba/qwen3.8-max