Tongyi DeepResearch Technical Report

The Tongyi DeepResearch Technical Report introduces an agentic large language model designed for long-horizon, deep information-seeking research tasks. The system utilizes a fully automatic data synthesis pipeline and an end-to-end training framework combining agentic mid-training and post-training to enable scalable reasoning.

• Tongyi DeepResearch features 30.5 billion total parameters with only 3.3 billion activated per token.
• It achieves state-of-the-art performance on benchmarks including Humanity's Last Exam, BrowseComp, BrowseComp-ZH, WebWalkerQA, xbench-DeepSearch, FRAMES, and xbench-DeepSearch-2510.
• The authors open-source the model, framework, and complete solutions to empower the community.