Several bots trade virtual money using well-known public strategies over the same period, so you can see which approaches would have held up. Every number comes from recorded simulations with slippage and costs included. This is a personal research tool and is not financial advice.
A public experiment where trading bots each follow a well-known strategy — momentum, mean reversion, value, options spreads and more — with $10,000 of virtual money, so you can compare which approach makes the most profit.
No. It is paper trading with virtual money, filled at real market prices with costs applied. It is research, not investment advice.
Yes, in a measured way. Every bot picks candidates with published rules. Three of them — Trend Breakout + AI, Mean Reversion + AI and Maestro + AI — then have an AI model (Claude, with DeepSeek as backup) review each pick before the open with its price history, fundamentals, recent headlines and next earnings date. The AI can approve, halve or reject a trade, never add one. Each runs next to its rules-only twin, so the gap shows what the AI added. They are tested live only, because an AI model has already seen past markets.
The bots publish a plan before each US market open, open positions are checked against live quotes about once an hour during the session, and everything is settled from the day's real prices the next morning (by 8:30 ET).
On the first trading day of every month each account is reset to $10,000 and the previous month is sealed into a monthly scoreboard, so results can be compared month by month. Since October 2026 the live accounts trade an aggressive profile: up to 4% of equity at risk per trade, positions up to 40% of the account and up to 5 at once.