Geometric (percentage-based) grid trading strategy designed for long-term operation on assets with bullish bias. Unlike arithmetic grids that use fixed dollar spacing, InfinityGrid uses percentage spacing so each completed grid cycle captures the same percentage return regardless of current price level. The grid has a configurable lower bound (capital protection floor) but no upper bound, continuing to generate sell orders at higher price levels indefinitely. Grid recycling creates an infinite buy-sell cycle: when a buy fills, a sell is placed one level up, and vice versa. Each completed cycle captures GRID_PERCENT% profit minus fees. State management uses a sliding window of active levels around current price, preventing state bloat while maintaining the infinite grid illusion. Optional profit compounding reinvests a portion of grid profits into larger level sizes.
Close to break-even at best in our backtests, with losses in falling markets. Try it on paper first.
Results from historical backtesting. Past performance does not guarantee future results.
Be the first to run InfinityGrid on OKX.
No backtests yet for this pair on OKX — try one below, right here.
Each bot's return on this version since it became stable (at most the last 30 days), averaged over bots with 5+ trades on it; drawdown marks open positions to their last stored price. Live bots only, else paper bots.
Create API keys on OKX and connect to TradeStaq.
Choose InfinityGrid from the strategy marketplace.
Set your trading pair, position size, and risk parameters.
Start the bot. It trades 24/7 based on the strategy rules.
Free calculator — set your risk per trade and stop-loss to get the exact position size before you deploy. No signup.
InfinityGrid uses to identify trading opportunities on OKX. It automatically places and manages orders based on the strategy's rules, with configurable risk parameters.
InfinityGrid pays 0.02%/0.05% maker/taker fees on OKX's futures market — factor that in when sizing frequent-trade runs.
InfinityGrid hasn't been backtested on OKX yet — run a free backtest below to see how it would have performed.