At the 2025 Global Developer Pioneer Conference, SenseTime announced the launch of the open source low-code platform LazyLLM. This innovative tool aims to lower the threshold for AI application development and allow more developers to easily participate in the application and innovation of AI technology. The design concept of LazyLLM is to allow developers to quickly build complex multi-agent large-scale applications through minimal code volume, thereby greatly improving development efficiency.
LazyLLM's name comes from the concept of "laziness", reflecting SenseTime's vision: allowing developers to easily achieve complex AI application development without having a deep programming background. Through this platform, developers can quickly combine various functional modules like building blocks, so as to focus on the implementation of business logic without being troubled by tedious technical details.
The core highlight of LazyLLM lies in its data flow-centric application development paradigm. This design allows developers to flexibly splice various functional modules and improve application performance by continuously optimizing data effects. In addition, the platform also provides one-click deployment functions, so developers can easily integrate intent recognition, knowledge base retrieval and big model capabilities to quickly implement products. After completion of development, the intelligent body can be seamlessly deployed to multiple platforms such as web pages, enterprise WeChat and DingTalk, greatly improving the usability and coverage of the application.
Through LazyLLM, developers can build a complete RAG multi-channel recall application in a very short time, support the construction and management of enterprise-level knowledge bases, and achieve efficient data processing and functional integration. This platform has also been widely recognized in terms of localization, solving the problem of insufficient adaptation of some foreign tools in China, and providing developers with a more convenient development experience.
In order to further lower the development threshold, SenseTime has also launched the Vientiane platform to provide one-stop AI application development solutions. Through visual drag and drop operations, users can build large-scale applications without writing code. The Vientiane platform supports a variety of functions, helping enterprises realize a closed loop from application development to iteration, greatly improving development efficiency, and providing strong support for enterprises' digital transformation.
Up to now, SenseTime has provided AI application solutions to multiple industries through LazyLLM and Vientiane platforms, helping more than 30 companies to improve their AI capabilities. These successful cases not only demonstrate SenseTime's technical strength in the field of AI, but also provide a strong proof of the transformation of AI technology into productivity.
Project address: https://github.com/LazyAGI/LazyLLM
Project documentation: docs.lazyllm.ai