Create Knowledge, not Data
Every day, millions of people ask AI the same questions.
"What causes aurora borealis?"
"How does mRNA vaccine work?"
"What is the history of the Silk Road?"
Each answer is brilliant. And each answer disappears the moment the chat window closes.
We are generating more knowledge than any civilization in history — and losing almost all of it.
Think about it: ChatGPT processes 1 billion+ queries per week. Claude, Gemini, Perplexity — billions more. The collective intelligence flowing through these conversations could fill libraries. Instead, it evaporates.
Wikipedia took 25 years to build 60 million articles with human editors. AI generates that volume of knowledge-grade content in days. But there is no place for it to land, accumulate, and grow.
This is the problem we are solving with GPTwiki.
The idea is simple
Ask any question. Get an AI-powered answer. That answer becomes a permanent, searchable wiki article — not a disposable chat message.
But here is what makes it different:
Multi-model freedom — Choose Claude, GPT-4o, or Gemini. Bring your own API key. No vendor lock-in.
Conversations that compound — Anyone can ask follow-up questions. Each answer enriches the article or spawns a new one.
15 languages, day one — Not translated. Natively generated. A Japanese user and a Brazilian user both get first-class content.
100% open source — MIT license. Self-host it. Fork it. Make it yours.
We have already seeded 100,000+ articles from Wikipedia across 15 languages to make sure the experience is rich from the first visit.
Why now?
The cost of generating high-quality knowledge just dropped to near zero. But the infrastructure to preserve and organize it has not caught up. Wikipedia is incredible but was designed for a pre-AI world — manual edits, lengthy review cycles, high contributor barriers.
We need a knowledge platform that is native to the AI era:
Instant creation through conversation
Community curation, not gatekeeping
Multi-model, multi-language by default
Open and self-hostable
What we are looking for
We are launching on Product Hunt soon and would love your support:
Makers — What features would you want in an AI-native encyclopedia? We are building in public and shipping fast.
Hunters — If this resonates with you, we would be honored to have you hunt us on launch day.
Everyone — Try it at gptwiki.net, break things, and tell us what you think.
The source code is at github.com/oratis/GPTwiki — stars are appreciated but honest feedback is worth more.
Let us stop letting knowledge evaporate. Let us build the encyclopedia the AI era deserves.
What do you think? Would you use something like this? What is missing?



Replies