MassGen

Run 15+ AI models in parallel — let them critique each other until they converge

MassGen is an open-source terminal-based multi-agent orchestration system that takes a fundamentally different approach to AI problem solving: instead of routing to a single model, it runs multiple frontier models (Claude, GPT, Gemini, Grok, and 12+ others) on the same task simultaneously. The agents can observe each other's outputs and iteratively critique and refine until they converge on a consensus answer. The tool features an interactive TUI with real-time visualization of parallel agent activity, MCP tool integration for connecting external capabilities, Docker-based code execution for safe sandboxing, and local model support via LM Studio and vLLM. It's particularly suited for complex coding tasks, research synthesis, and decisions where you want multiple perspectives rather than trusting a single model's confident answer. Released in early April 2026 under Apache 2.0, MassGen fills a gap between single-agent tools and expensive enterprise orchestration platforms. The "ensemble" approach mirrors how expert panels work — divergent perspectives followed by structured critique — and the terminal-native UX keeps it close to developer workflows without requiring a new cloud subscription.

Panel Reviews

The Builder

Developer Perspective

Ship

“The terminal-native ensemble approach is genuinely novel. Being able to spin up Claude, GPT-5, and Gemini on the same hard problem and watch them debate is something I've wanted for ages. Adds real value for decisions where a single model's confident wrong answer would cost you hours.”

The Skeptic

Reality Check

Skip

“Running 15 models in parallel means paying API costs for all of them, which adds up fast. And 'convergence by critique' is speculative — models may just agree with each other's mistakes rather than catch them. I'd want hard benchmark evidence before trusting ensemble output over a single well-prompted Opus call.”

The Futurist

Big Picture

Ship

“Single-model pipelines have hit their ceiling on complex tasks; ensemble approaches that leverage model diversity are the next frontier. MassGen makes this accessible at the terminal level before it becomes a $50k enterprise feature from AWS.”

The Creator

Content & Design

Ship

“For creative tasks like copywriting, script outlines, or design brief generation, having multiple AI voices critique each other produces far more interesting outputs than any single model. The parallel TUI visualization is genuinely addictive to watch in action.”