Anfeng Li
Anfeng Li is a researcher whose AI work spans large language models and model evaluation. In 2025, he coauthored the Qwen3 Technical Report, which describes dense and mixture-of-experts models, a unified thinking framework, support for 119 languages and dialects, and Apache 2.0 access. In 2026, he contributed to Qwen3.5-Omni and coauthored HLE-Verified, a systematic revision of Humanity's Last Exam with a transparent, component-wise protocol and a fine-grained error taxonomy. As of October 2026, he works on Alibaba's Qwen Team.
Timeline
-
May 2025
The Qwen3 report
He coauthors the Qwen3 Technical Report in 2025. It describes dense and mixture-of-experts models with parameter scales from 0.6 to 235 billion.
-
May 2025
Combining reasoning modes
The 2025 Qwen3 model combines thinking and non-thinking modes in one framework.
-
May 2025
Expanding language coverage
Its 2025 report expands support from 29 to 119 languages and dialects, and makes all Qwen3 models publicly accessible under Apache 2.0.
-
February 2026
Revising Humanity's Last Exam
He coauthors HLE-Verified in 2026, a systematic revision of Humanity's Last Exam with a transparent, component-wise verification protocol and fine-grained error taxonomy.
-
April 2026
Contributing to Qwen3.5-Omni
He contributes to the Qwen3.5-Omni Technical Report in 2026.
-
As of October 2026
On Alibaba's Qwen Team
As of October 2026, he works on Alibaba's Qwen Team.
Most-cited works
Sources and credits
Photo: We have no published photo of Anfeng Li yet. Send us a free-licensed photo.
- [2505.09388] Qwen3 Technical Report
- GitHub - QwenLM/Qwen3: Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud. · GitHub
- Paper page - Qwen3 Technical Report
- [2602.13964] HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam
- HLE-Verified: A Systematic Verification and Structured Revision of Humanity’s Last Exam
- Qwen3.5-Omni Technical Report

