Claude turns $2,000 to $12,000 overnight on Polymarket; Here is how
Claude turns $2,000 to $12,000 overnight on Polymarket; Here is how 
Cryptocurrency Apr 8, 2026 Share
A trader successfully leveraged Claude AI to make a $10,000 profit on Polymarket with $2,000 in capital in one night.
In a 17-second snippet analyzed by Finbold on April 8, the Anthropic-backed chatbot helped developer Mary Evan create a custom interface for Polymarket, a blockchain-based prediction market. As such, the trader earned approximately 500% on the investment.
Claude-powered Polymarket trading terminal. Source: Mary Evan
The Claude-powered bot traded different events, including in crypto, finance, and technology. Throughout its trading cycle, the bot scanned Polymarket for mispriced markets, extracted arbitrage opportunities, and traced wallets that copied these strategies.
“Claude created a monitoring terminal and copy-traded found wallets using a Telegram copy-trading bot. It’s not a script and not even the bot; it’s an AI agent that is improving with each found wallet,” Evan stated.
Can Claude repeat its $10,000 Polymarket win, or was it luck?
The number of instances in which Claude has helped traders make huge profits has been increasing in recent months. On Tuesday, Finbold reported that a Claude AI-powered arbitrage bot had turned $600 into $10,000 in about 2 days.
Last week, Finbold noted that a Claude-powered trading bot focused on arbitrage sports bets had generated a profit of approximately 330 million percent in eight months. With Evan’s case of turning $2,000 into $12,000 in 24 hours, the case for using AI to develop autonomous trading algorithms is strengthened.
In another study, a trading bot that leverages copy trading on Polymarket grew an account balance of $0.12 into more than $526,000 at press time, making more than 17,000 predictions.
As such, it could be possible for anyone to prompt Claude to instantly reverse-engineer any trading bot strategy from a GitHub repository. However, caution should be exercised while building autonomous trading algorithms, as a small error can cause huge losses, as Finbold previously highlighted.
