At the 2025 Global Developer Pioneer Conference, SenseTime announced the launch of its open-source low-code platform, LazyLLM, aiming to lower the barrier to entry for AI application development. Developers can now build complex multi-agent large model applications with approximately 10 lines of code.
LazyLLM not only helps users rapidly develop AI applications but also simplifies the development process, making it as easy and enjoyable as building with blocks.
LazyLLM's core strength lies in its dataflow-centric application development paradigm, supporting various concatenation methods to continuously improve data efficiency. The platform also features one-click deployment, allowing developers to quickly combine intent recognition, knowledge base retrieval, and large model capabilities for rapid product deployment. Once developed, agents can be easily deployed to multiple platforms, including web pages, WeChat Work, and DingTalk.
With LazyLLM, developers can quickly build complete RAG multi-path retrieval applications, supporting the construction and management of enterprise-level knowledge bases, and enabling efficient data processing and functional integration. Furthermore, LazyLLM's localization features have been widely praised by developers, addressing the shortcomings of some foreign tools.
To further lower the development threshold, SenseTime has also launched the Wanxiang platform, providing a one-stop AI application development solution. Users can build large model applications through visual drag-and-drop operations, requiring no coding at all in some cases. The Wanxiang platform supports various functions, helping enterprises achieve a closed loop from application development to iteration, significantly improving development efficiency.
To date, SenseTime has leveraged LazyLLM and the Wanxiang platform to provide AI application solutions for multiple industries, helping over 30 companies enhance their AI capabilities and transforming AI into productive force.
Project Address: https://github.com/LazyAGI/LazyLLM
Project Documentation: docs.lazyllm.ai
