AI Disruption

AI Disruption

Tongyi DeepResearch Tops Open-Source Agent Rankings

Alibaba's Tongyi DeepResearch tops open-source AI agent rankings—outperforming rivals like OpenAI with its 34B lightweight model. Fully open-sourced on GitHub & Hugging Face.

Meng Li's avatar
Meng Li
Sep 18, 2025
∙ Paid
9
5
Share

"AI Disruption" Publication 7700 Subscriptions 20% Discount Offer Link.


Alibaba Releases Tongyi DeepResearch: A 30B-Parameter Open-Source Agentic  LLM Optimized for Long-Horizon Research - MarkTechPost

Alibaba Open Source Launches Its First Deep Research Agent Model - Tongyi DeepResearch!

Compared to ReAct Agents based on foundation models and closed-source Deep Research Agents, this 30B-A3B lightweight model, Tongyi DeepResearch, achieves SOTA performance on multiple authoritative evaluation benchmarks, including HLE, BrowseComp-zh, and GAIA, surpassing Agent models like OpenAI Deep Research and DeepSeek-V3.1.

Currently, Tongyi DeepResearch's model, framework, and solutions are all fully open-sourced. Interested friends can download the model and code from GitHub, Hugging Face, and the ModelScope community.

Let's look at the specific details of the model below.

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2025 Meng Li
Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture