Zero-code multi-agent orchestration platform from Tsinghua University for developing everything — from software to data visualization and deep research — using LLM-powered agent collaboration.
An open-source zero-code platform where AI agents collaborate to develop software, analyze data, generate 3D content, and conduct deep research — evolved from a virtual software company into a general multi-agent orchestration system.
ChatDev has evolved from a specialized software development multi-agent system into a comprehensive multi-agent orchestration platform. In January 2026, the team released ChatDev 2.0 (DevAll) — a zero-code multi-agent platform for 'Developing Everything' — moving beyond the original virtual software company concept.
ChatDev 2.0 empowers users to rapidly build and execute customized multi-agent systems through simple configuration files. No coding is required — users define agents, workflows, and tasks to orchestrate complex scenarios including software development, data visualization, 3D generation, deep research, and more. The classic ChatDev 1.0 with its CEO/CTO/Programmer/Tester role-playing paradigm has been moved to a legacy branch for maintenance.
The original ChatDev 1.0 simulated a virtual software company where AI agents collaborated through chat-based interactions. It assigned agents to software development roles (CEO, CTO, Programmer, Tester, Art Designer) and organized development into phases — Designing, Coding, Testing, and Documenting — with each phase involving specific agent pairs communicating through a chat chain. This conversational approach created transparent development dialogues and included the distinctive 'inception prompting' technique for role assignment.
ChatDev 2.0 generalizes this into a flexible multi-agent orchestration layer. It supports the puppeteer-style paradigm for multi-agent collaboration introduced in their NeurIPS 2025 paper, where a learnable central orchestrator optimized with reinforcement learning dynamically activates and sequences agents to construct efficient, context-aware reasoning paths. The platform also incorporates MacNet (Multi-Agent Collaboration Networks) using directed acyclic graphs for effective task-oriented collaboration across various topologies and among more than a thousand agents without exceeding context limits.
The platform supports multiple LLM providers including OpenAI, Anthropic, and local models via Ollama. The experience pool feature stores solutions and patterns from previous sessions, enabling agents to learn from past projects and apply proven solutions to new tasks.
Honest assessment: ChatDev 2.0 represents a significant leap from the original research demo. The zero-code orchestration approach makes multi-agent systems accessible beyond developers, and the research backing (NeurIPS 2025 acceptance) gives credibility to the collaboration paradigms. However, it remains an academic project with less production polish than commercial alternatives like CrewAI or AutoGen. Output quality varies with the underlying LLM and task complexity. It's strongest as a research platform for exploring multi-agent patterns and for rapid prototyping, but teams needing production reliability should plan for significant customization and human review of outputs.
Was this helpful?
ChatDev 2.0 transforms the original virtual software company concept into a general-purpose zero-code multi-agent orchestration platform. Backed by serious research (NeurIPS 2025), it offers unique capabilities like puppeteer orchestration and MacNet scaling. Best for researchers and experimenters exploring multi-agent paradigms; not yet a production-grade alternative to commercial frameworks.
Define and execute customized multi-agent systems through simple configuration without writing code. Supports diverse scenarios including software development, data visualization, 3D generation, and deep research tasks.
Use Case:
Setting up a multi-agent research pipeline that gathers data, analyzes trends, generates visualizations, and produces a formatted report — all defined through configuration files rather than code.
A learnable central orchestrator optimized with reinforcement learning that dynamically activates and sequences agents to construct efficient, context-aware reasoning paths. Published and accepted at NeurIPS 2025.
Use Case:
Complex problem-solving where the optimal agent sequence isn't known in advance — the orchestrator learns which agents to activate and in what order based on the task context.
Multi-Agent Collaboration Networks using directed acyclic graphs for task-oriented collaboration across various topologies and among more than 1,000 agents without exceeding LLM context limits.
Use Case:
Scaling multi-agent collaboration to handle complex tasks requiring dozens of specialized agents working in parallel and serial arrangements across different subtasks.
Complete simulation of a software company with AI agents playing CEO, CTO, Programmer, Tester, and Art Designer roles, collaborating through natural language chat chains across Design, Coding, Testing, and Documentation phases.
Use Case:
Generating complete software prototypes from natural language descriptions, with transparent development dialogues showing every design decision and code change rationale.
Knowledge base that stores solutions, patterns, and fixes from previous sessions. Incorporates Iterative Experience Refinement (IER) where agents enhance shortcut-oriented experiences to efficiently adapt to new tasks.
Use Case:
Running multiple code generation sessions where the system improves over time — authentication patterns solved in project A are automatically applied when similar patterns appear in project B.
Configure organizational structures, role definitions, development phases, workflow rules, and collaboration topologies. Supports chain, graph, and network-based agent communication patterns.
Use Case:
Creating a data science-focused agent team with ML Engineer, Data Analyst, and DevOps roles instead of traditional software development roles, with custom workflow phases.
Free
forever
Ready to get started with ChatDev?
View Pricing Options →Exploring different agent collaboration topologies (chain, graph, network), orchestration strategies, and role configurations for academic research and practical system design.
Generating complete software prototypes with functional code, documentation, and visual mockups from natural language descriptions — useful for validating ideas before committing developer resources.
Orchestrating multi-agent workflows that gather data, perform analysis, generate visualizations, and produce formatted reports using ChatDev 2.0's zero-code configuration.
Teaching software engineering processes through realistic multi-agent collaboration that demonstrates real-world development workflows, decision-making, and team dynamics.
ChatDev works with these platforms and services:
We believe in transparent reviews. Here's what ChatDev doesn't handle well:
ChatDev 1.0 was specifically a virtual software company simulation with fixed roles (CEO, CTO, Programmer, etc.). ChatDev 2.0 (DevAll), released January 2026, is a general-purpose zero-code multi-agent orchestration platform that can handle software development, data visualization, 3D generation, deep research, and more. The 1.0 version is maintained on a legacy branch.
A typical software project generation costs $0.20-$2.00 in LLM API calls with GPT-4. The conversational approach may use more tokens than structured frameworks due to multi-turn agent discussions. Using local models via Ollama eliminates API costs entirely, though output quality depends on the model.
ChatDev is best suited for prototyping, research, and initial code generation rather than production development. The generated code serves as a starting point that human developers should review, test, and refine. ChatDev 2.0's broader capabilities make it more practical for data analysis and research workflows.
ChatDev 1.0 primarily generates Python applications with optimized testing and debugging workflows. ChatDev 2.0 is more flexible since it orchestrates general-purpose agents, but Python remains the best-supported language. Community forks have added better support for JavaScript and TypeScript.
ChatDev is more research-oriented with unique contributions like MacNet collaboration networks and puppeteer orchestration (NeurIPS 2025). CrewAI and AutoGen offer more production-ready features and broader ecosystem integrations. Choose ChatDev for exploring novel multi-agent paradigms; choose CrewAI or AutoGen for production deployments.
Weekly insights on the latest AI tools, features, and trends delivered to your inbox.
In January 2026, ChatDev released version 2.0 (DevAll) — a zero-code multi-agent orchestration platform replacing the original software company simulation. New capabilities include puppeteer-style orchestration with reinforcement learning (accepted at NeurIPS 2025), MacNet collaboration networks supporting 1,000+ agents, and expanded use cases beyond code generation to data visualization, 3D generation, and deep research.
People who use this tool also find these helpful
Open-source multi-agent framework forked from Microsoft AutoGen, using conversation-driven coordination to orchestrate AI agents for code generation, research, and collaborative problem-solving.
The next-generation AG2 platform with AgentOS runtime, framework interoperability, teachable agents, and enhanced planning for production multi-agent systems.
Open-source Python framework that organizes AI agents into company-like hierarchies with strict communication channels. Built on the OpenAI Agents SDK. Free to use; you pay only for API calls to the LLM providers.
Research-first multi-agent framework with #1 GAIA benchmark performance, designed for studying agent societies and role-playing simulations at scale
Open-source agent framework built on Llama models with local deployment options and community-driven development.
Multi-agent software company simulation platform.
See how ChatDev compares to CrewAI and other alternatives
View Full Comparison →AI Agent Builders
CrewAI is an open-source Python framework for orchestrating autonomous AI agents that collaborate as a team to accomplish complex tasks. You define agents with specific roles, goals, and tools, then organize them into crews with defined workflows. Agents can delegate work to each other, share context, and execute multi-step processes like market research, content creation, or data analysis. CrewAI supports sequential and parallel task execution, integrates with popular LLMs, and provides memory systems for agent learning. It's one of the most popular multi-agent frameworks with a large community and extensive documentation.
Agent Frameworks
Open-source multi-agent framework from Microsoft Research with asynchronous architecture, AutoGen Studio GUI, and OpenTelemetry observability. Now part of the unified Microsoft Agent Framework alongside Semantic Kernel.
AI Agent Builders
Graph-based stateful orchestration runtime for agent loops.
Multi-Agent Builders
Multi-agent software company simulation platform.
No reviews yet. Be the first to share your experience!
Get started with ChatDev and see if it's the right fit for your needs.
Get Started →Take our 60-second quiz to get personalized tool recommendations
Find Your Perfect AI Stack →Explore 20 ready-to-deploy AI agent templates for sales, support, dev, research, and operations.
Browse Agent Templates →