ChatDev
ChatDev is a multi-agent collaborative software development framework based on large language models (LLMs). It simulates the organizational structure of a software company by assigning development tasks to AI agents playing distinct roles—such as CEO, programmer, and tester—which communicate and cooperate to automatically complete the full software development pipeline, from requirements analysis through code generation, testing, and documentation writing.
Core Facts
Timeline (last 90 days)
ChatDev, developed by a domestic team, virtualizes the entire software development process, using AI group collaboration to achieve results stronger than a single agent.
ChatDev, developed by a domestic team, virtualizes the entire software development process and achieves more powerful results than a single agent through AI group collaboration.
ChatDev simulates the organizational structure of a virtual software company, setting up role-based Agents such as CEO, CTO, programmer, and test engineer, and driving the development pipeline through role-playing dialogue.
ChatDev uses a Chat Chain to allow roles to collaboratively complete the entire process from requirements analysis to code testing in sequence.
The role-playing plus phased mechanism effectively reduces the risks brought by single-model hallucination and context drift.
ChatDev is a Python-based open-source project in which developers can customize agent roles, dialogue workflows, and collaboration rules.
Compared to earlier versions, ChatDev 2.0 expands from conversational software development to a broader range of development tasks, emphasizing the use of multi-agent collaboration to cover all development scenarios.
ChatDev is a multi-agent collaboration framework open-sourced by the OpenBMB team, with the goal of completing the entire software development process through multiple AI agents driven by large language models.
ChatDev adopts a division-of-labor collaboration approach, breaking down the entire development process into multiple stages, each handled by agents playing specific roles.
ChatDev's design follows the classic software development waterfall model philosophy, with roles exchanging information through natural language dialogue.
1 more timeline events
All Facts (12)
ChatDev implements an automated software development pipeline from requirement description to runnable code through role-playing (with AI roles such as CEO, CTO, and programmer collaborating)
65%UnverifiedCompared to earlier versions, ChatDev 2.0 expands from conversational software development to a broader range of development tasks, emphasizing the use of multi-agent collaboration to cover all development scenarios.
60%UnverifiedChatDev, developed by a domestic team, virtualizes the entire software development process, using AI group collaboration to achieve results stronger than a single agent.
50%UnverifiedChatDev, developed by a domestic team, virtualizes the entire software development process and achieves more powerful results than a single agent through AI group collaboration.
50%UnverifiedChatDev simulates the organizational structure of a virtual software company, setting up role-based Agents such as CEO, CTO, programmer, and test engineer, and driving the development pipeline through role-playing dialogue.
50%UnverifiedChatDev uses a Chat Chain to allow roles to collaboratively complete the entire process from requirements analysis to code testing in sequence.
50%UnverifiedChatDev is a multi-agent collaboration framework open-sourced by the OpenBMB team, with the goal of completing the entire software development process through multiple AI agents driven by large language models.
50%UnverifiedChatDev is a Python-based open-source project in which developers can customize agent roles, dialogue workflows, and collaboration rules.
50%UnverifiedChatDev adopts a division-of-labor collaboration approach, breaking down the entire development process into multiple stages, each handled by agents playing specific roles.
50%UnverifiedChatDev's design follows the classic software development waterfall model philosophy, with roles exchanging information through natural language dialogue.
50%UnverifiedThe role-playing plus phased mechanism effectively reduces the risks brought by single-model hallucination and context drift.
50%UnverifiedIn 2024, Tsinghua University launched the ChatDev project, which uses multiple AI agents to simulate a software company completing a development workflow.
50%