Qwen3.8-27B
Reasoning modelCoding modelQwen3.8-27B
Qwen3.8-27B is an AI model published by Alibaba, released on 2026-08-14, for Reasoning model, with 27B parameters, and 256K context length, with a 94.60 score on MathVision.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Model basics
Open source & experience
Official resources
API details
Benchmark Results
Qwen3.8-27B currently shows benchmark results led by LiveCodeBench (6 / 124, score 90.30), OSWorld-Verified (3 / 26, score 84.30), CharXiv RQ (2 / 15, score 90.20). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.
Coding and Software Engineer
4 evaluationsAI Agent - Tool Usage
2 evaluationsMultimodal Understanding
7 evaluationsCompare with other models
- Peer modelQwen3.8-27B vs DeepSeek-V4-Flash8 benchmarks
- Peer modelQwen3.8-27B vs Claude Sonnet 55 benchmarks
- Earlier versionQwen3.8-27B vs Qwen3.5-27B4 benchmarks
- Earlier versionQwen3.8-27B vs Qwen3.6-27B4 benchmarks
- Peer modelQwen3.8-27B vs GPT-5.6 Terra4 benchmarks
- Peer modelQwen3.8-27B vs Gemini 3.7 Flash4 benchmarks
Want a custom combination? Open the compare tool
Publisher
Model Overview
Qwen3.8-27B is an AI model published by Alibaba, released on 2026-08-14, for Reasoning model, with 27B parameters, and 256K context length, with a 94.60 score on MathVision.
DataLearner on WeChat
Follow DataLearner on WeChat for AI model updates and research notes.
