Five AI minds · one question
Project Argon 5
Five different models answer independently, read each other, challenge the reasoning — and try to convince you.No easy answers.
GeminiChatGPTClaudeDeepSeekQwenThis is not five chatbots answering in parallel. It is one board: they read each other's reasoning, name what is strong and what is weak, defend their position or abandon it — and they change their minds when the evidence says they should.
The structure
Four rounds, then you choose
Round 1
Independent answers
All five answer without seeing each other. Five genuinely independent starting positions.
Round 2
The board reads the board
Each member reads the other four, names the strongest argument on the table and attacks the weakest.
Round 3
The board questions you
The board turns on you. Each member asks the one question that would change their mind.
Round 4
Closing arguments
Closing arguments. You decide who argued it best.
Then
You decide
Pick the member who convinced you. The conversation is saved — return to it whenever you like.
Every session is exactly 20 model calls. No endless tail, no surprise bill.
The board
Five minds, five providers
via Google
via OpenAI
via Anthropic
via DeepSeek
via Groq
The titles are identities, not positions. Every model is free to disagree with what its title suggests — and to say plainly when another member argued it better.
Pricing
Same structure, sharper minds
Standard
1 per day, then 0,15 € per session from your balance
Five efficient models. One free session per day, then from your balance.
Medium
per session, paid from your balance
A stronger mix of models for harder questions.
High
per session, paid from your balance
Five flagship models. The sharpest debate available.
All prices include VAT. Every tier runs the same 20-call structure — only the models differ.