Skip to main content
PaperLLMsv2.5

Qwen2.5 Technical Report

by Alibaba Cloud / Qwen Team · open-source · Last verified 2026-03-17

Qwen2.5 is a comprehensive family of open-source LLMs (0.5B to 72B parameters) trained on 18 trillion tokens including significantly expanded coding and mathematics data, achieving state-of-the-art open-source performance on coding (HumanEval), mathematics (MATH), and multilingual benchmarks. The series includes specialized Qwen2.5-Coder and Qwen2.5-Math variants.

https://arxiv.org/abs/2412.15115
B
BAbove Average
Adoption: AQuality: A+Freshness: A+Citations: B+Engagement: F

Specifications

License
Qwen License / Apache 2.0 (sub-72B)
Pricing
open-source
Capabilities
text-generation, code-generation, math-reasoning, multilingual, instruction-following, long-context
Integrations
Hugging Face, Ollama, vLLM, Together AI, Alibaba Cloud
Use Cases
code-generation, math-problem-solving, multilingual-qa, instruction-following, agent-tasks
API Available
Yes
Tags
llm, qwen, alibaba, open-source, multilingual, 2024, coding
Added
2026-03-17
Completeness
100%

Index Score

69
Adoption
80
Quality
91
Freshness
92
Citations
75
Engagement
0

Put AI to work for your business

Deploy this paper alongside autonomous AaaS agents that handle tasks end-to-end — no babysitting required.

Explore the full AI ecosystem on Agents as a Service