Tongyi DeepResearch Technical Report

TL;DR

Tongyi DeepResearch is a large language model designed for complex, long-horizon research tasks, developed through an end-to-end training framework with automatic data synthesis, achieving state-of-the-art results and open-sourcing its resources.

Contribution

The paper introduces a novel agentic large language model with a scalable automatic training pipeline for deep research tasks, setting new performance benchmarks.

Findings

Achieves state-of-the-art performance on multiple deep research benchmarks.

Uses a highly scalable, fully automatic data synthesis pipeline.

Features a 30.5 billion parameter model with efficient activation.

Abstract

We present Tongyi DeepResearch, an agentic large language model, which is specifically designed for long-horizon, deep information-seeking research tasks. To incentivize autonomous deep research agency, Tongyi DeepResearch is developed through an end-to-end training framework that combines agentic mid-training and agentic post-training, enabling scalable reasoning and information seeking across complex tasks. We design a highly scalable data synthesis pipeline that is fully automatic, without relying on costly human annotation, and empowers all training stages. By constructing customized environments for each stage, our system enables stable and consistent interactions throughout. Tongyi DeepResearch, featuring 30.5 billion total parameters, with only 3.3 billion activated per token, achieves state-of-the-art performance across a range of agentic deep research benchmarks, including Humanity's Last Exam, BrowseComp, BrowseComp-ZH, WebWalkerQA, xbench-DeepSearch, FRAMES and xbench-DeepSearch-2510. We open-source the model, framework, and complete solutions to empower the community.