BettaFish — also called WeiYu — is a free, self-hosted tool that watches over 30 social media platforms for you, then produces a structured research report on any topic you ask about in plain language.
You type a question. It goes out, gathers data from across the internet, and comes back with charts, analysis, and conclusions. Think of it as hiring a small research team that never sleeps and doesn't invoice you.
Most AI tools give you one answer. BettaFish does something different: it spins up several specialized agents — one to search, one to read, one to find patterns, one to write — and before they hand you anything, they argue with each other. A layer called the ForumEngine makes the agents challenge each other's conclusions, flag weak evidence, and push back. Only then does the final report get written.
That debate step is genuinely unusual. It's closer to how a good analyst actually works than to how most AI tools behave.
A 20-year-old built this as a university project, with no pre-made AI frameworks — just from scratch. It hit the top of GitHub trending with over 4,000 stars in a single day.
If you've ever paid for a brand monitoring service, or spent hours manually checking what people are saying about your business online, this is the kind of tool worth bookmarking. It won't replace your judgment — but it can do the legwork.
Self-hosted — You run the tool on your own computer or server, not someone else's. Your data stays yours.
Agent — An AI that doesn't just answer a question but takes a series of steps to complete a task, like a researcher who goes off and actually does the work.
Multi-agent system — Several AI agents working together, each with a different job, like a small team rather than a single assistant.
Open-source — The code is public and free. Anyone can use it, inspect it, or improve it.