A team of developers at AI dev platform Hugging Face, led by co-founder and chief scientist Thomas Wolf, has successfully built an open-source version of OpenAI's deep research tool. Dubbed Open Deep Research, the project aims to provide a more accessible and affordable alternative to OpenAI's proprietary technology, which is currently only available to users subscribed to the $200-a-month ChatGPT Pro plan.

OpenAI's deep research tool, unveiled during an event on Sunday, is a powerful technology that crawls the web to compile research reports on any subject. While impressive, its limited availability has sparked concerns about accessibility and fairness. The Hugging Face team's project seeks to address these concerns by creating an open-source framework that can be used by anyone, regardless of their financial means.

The Open Deep Research project consists of an AI model, OpenAI's o1, and an open-source "agentic framework" that helps the model plan its analysis and guides it to use tools like search engines. Although o1 is a proprietary model, the team claims it delivered better performance than "open" models such as DeepSeek's R1. In less than 24 hours, the researchers were able to harness o1 to use a simple, text-based browser and a "text inspector" toolkit to read files across the web.

Open Deep Research can navigate the web autonomously, scrolling through pages, manipulating files, and even running calculations with data. On GAIA, a benchmark for general AI assistants, Open Deep Research achieves a score of 54%, compared to OpenAI deep research's score of 67.36%. While the results are promising, the team acknowledges that there is still room for improvement and has made the source code available on GitHub for inspection and feedback.

It's worth noting that Open Deep Research is not the only open-source alternative to OpenAI's deep research tool. There are several other reproductions available online, some of which rely on open models and tooling. However, the crucial component they lack is o3, the model underpinning deep research. Few, if any, models beat o3 on benchmarks related to answering complex questions and information gathering, which means that deep research alternatives may not quite measure up to the real thing.

Despite these limitations, the Open Deep Research project is a significant step towards democratizing access to advanced AI technologies. By making the source code available, the Hugging Face team is inviting the developer community to contribute to the project and help improve its performance. As the project continues to evolve, it may potentially challenge OpenAI's dominance in the field and pave the way for more innovative and accessible AI solutions.

In conclusion, the Open Deep Research project is a testament to the power of open-source collaboration and innovation. By creating an open-source alternative to OpenAI's deep research tool, the Hugging Face team is promoting greater accessibility, fairness, and transparency in the development of AI technologies. As the project continues to grow and evolve, it may have far-reaching implications for the future of AI research and development.