Simulate a crowd of AI characters reacting to a piece of news

Made by BaiFu

A Python engine that takes a starting piece of real information — a news item, a policy draft, a financial signal — and runs a simulated population of AI agents against it.

Each agent is given its own personality, a memory that persists, and rules for how it behaves, and the project’s idea is that watching them interact suggests how a real crowd might respond. The documentation is mostly in Chinese.

Can I use this?

Runs in a terminal · needs an AI model key

You'll need Python, a terminal and an AI model key · Setup needed

Worth knowing A simulation of opinion is not a measurement of it, and the project offers no evidence that its predictions match real outcomes. Running thousands of agents means thousands of model calls, which costs money. It is AGPL-3.0, so a hosted modified version must publish its source, and most of the documentation is in Chinese. BuildTube has not run or verified this software.

Why it's here

Not trending this week

Last seen on the GitHub trending list on 20 September 2026. It stays on BuildTube so you can still find it.

76,720 people have starred it on GitHub.

Numbers checked on 6 October 2026; not refreshed since.

Behind it

See the code on GitHub

BuildTube has not run or verified this project. Everything above is written from what the creator published.

Made this?