LLM Reference

Qwen3.8-27B

Released
2026-08-17
Last refreshed
2026-08-24
Status
Researched 1d ago
Open sourceCommercial use: permittedMultimodalCodingRAGAgentsLong contextVisionJSON / Tool use

Qwen3.8-27B is worth evaluating for coding, rag, and agents when its provider route and context window match the workload.

Use it for

  • Teams evaluating coding, rag, and agents
  • Workloads that can use a 1m context window
  • Buyers comparing 1 tracked provider route

Do not use it for

  • Workloads where another current model has stronger sourced task evidence
Specifications
Family
Qwen3.8
Released
2026-08-17
Context
1m
Max output
131,072
Parameters
27B
Architecture
Decoder Only
Openness
Open source
License
Apache 2.0OSI-approvedCommercial use: permitted
Weights
Available
Code
Unknown
Created by

AI research institute of Alibaba Group.

Hangzhou, Zhejiang, China
Founded 2017
Website
Pricing
Output / 1M
$3.00
Input / 1M
$0.500

Cheapest of 1 route · Alibaba Cloud PAI-EAS · cache read $0.050

About

Qwen3.8-27B is the Qwen3.8 generation's native vision-language dense model, building on Qwen3.6-27B with key improvements in coding and office productivity across text and visual modalities. It enables more reliable end-to-end completion of complex tasks with consistently trustworthy results. Apache 2.0 weights are on Hugging Face; the hosted qwen3.8-27b API supports 1M context, reasoning, function calling, tool use, structured outputs, and prompt caching.

Top use-case fit: coding, agents, and build tasks

Coding

Q/$ D

2 relevant benchmarks in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Agents

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare API pricing across 1 providers for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MCacheRoute
Alibaba Cloud PAI-EAS$0.500$3.00read $0.050
Serverless

Capabilities

VisionMultimodalReasoningJSON / Tool useStructured OutputsPrompt Caching

Benchmark peer barsfor Coding

Benchmark scores(6)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionEvaluationSource
Terminal-Bench 2.173.0Terminal Bench 2.1 (Terminus)Observed 2026-08-14
Harness: Terminus
Source
SWE-bench Pro61.7SWE-bench ProObserved 2026-08-14Source
DeepSWE 1.142.2DeepSWE 1.1Observed 2026-08-14Source
Google-Proof Q&A89.2GPQA DiamondObserved 2026-08-14Source
LiveCodeBench90.3LiveCodeBench v6Observed 2026-08-14Source
OSWorld-Verified84.3OSWorld-VerifiedObserved 2026-08-14Source

Migration checks

No linked migration route is available for this model yet.

API versions

qwen3.8-27b

Rankings & picks(1)