In simple English
Tongyi Deep Research is an open-source research assistant that searches the web and reads documents to answer complex questions. You can run it on your own computer or call it through an API. It breaks a big question into smaller search steps. Then it reads pages, gathers facts and writes a final report. It costs you computer time, API fees for searches, or external hosting costs.
What it is
This repository provides scripts and weights for Tongyi DeepResearch, an agentic language model with 30.5 billion total parameters and 3.3 billion active parameters per token. It uses a ReAct inference loop to search the web, read web pages and parse uploaded files.
What it is good for
Developers and researchers who want to build custom automated research bots that can answer complex multi-step questions without manual searching.
- Open-source model weights available on HuggingFace and ModelScope.
- Supports local execution or cloud calls via OpenRouter.
- Performs multi-step web searches and document parsing automatically.
- Requires Python 3.10 and a dedicated virtual environment.
- Needs external API keys for web search and page reading services.
- Running locally demands significant GPU memory for the 30B model.
Manual web searching and multi-tab browser research
OpenAI API or OpenRouter
- Significant-Gravitas/AutoGPT188k stars
Another autonomous agent framework that performs multi-step tasks using language models.
- assafelovic/gpt-researcher30k stars
An open-source autonomous agent designed specifically for online research and report writing.
- ggml-org/llama.cpp132k stars
A C and C++ library and command-line tool for running language models locally.
- EleutherAI/lm-evaluation-harness14k stars
Runs standard question sets against a language model and reports the scores.
- mudler/LocalAI50k stars
Serves text, vision, speech, image and video models from one self-hosted service that imitates the paid interfaces.
conda create -n react_infer_env python=3.10.0 conda activate react_infer_env pip install -r requirements.txt lp .env.example .env