Build ML-inspired crypto strategies without code. Coinrule turns your signals into automated entries, exits and risk controls across major exchanges in minutes.
410k+
users building rules
1M+
strategies created
20+
Exchanges Supported
Most traders can spot patterns but struggle to execute consistently. With Machine learning trading, you can translate a model signal into clear if-then rules on Coinrule and let automation handle the clicks. For example, if BTC drops 4% in 6 hours and RSI(1H) is below 30, buy 1% of balance, then place a 2.5% stop loss and a 6% take profit. Connect to Binance, Coinbase, Kraken or Bybit via API keys while keeping funds on the exchange.
Instead of deploying notebooks and servers, you can express your logic as conditions, indicators and schedules. Import a signal as a webhook alert, then combine it with filters like trend direction, volume thresholds or volatility regimes. This keeps the workflow practical: research elsewhere, execute here. You can clone rules, adjust parameters, and keep a clean audit trail of every change.
Traders often overfit when they only optimize for returns. Coinrule helps you test ideas with realistic constraints like position sizing, cooldown timers and max open trades. Try a simple guardrail: stop trading for 12 hours after 3 consecutive losses, or reduce order size from 2% to 0.8% when 24-hour volatility exceeds 6%. These controls make systematic execution more robust when market conditions shift.
Signal In, Rules Out
When your model triggers, Coinrule can react instantly with a predefined playbook. Route the same logic to multiple venues, or keep it isolated to one exchange for cleaner evaluation. Add confirmation like MA(50) above MA(200) before entering, and require volume to be 20% above its 14-day average to avoid thin moves. You get speed without sacrificing discipline.
Guardrails First
Every strategy should include exits and limits. Use hard stops, trailing take profit, max daily loss, and exposure caps per coin so one bad regime does not dominate results. If a position reaches +8%, you might tighten the trailing stop from 4% to 2% to protect gains. These rules keep the system stable even when signals are noisy.
Volatility Regime Switcher
Trade only when volatility fits your regime. Enter on ATR expansion breakouts and reduce size or pause when ATR compresses. Exit with a time stop if momentum fades.
Mean Reversion with Z-Score
Buy when price falls below a moving average by a set Z-score. Sell on mean reversion and cut losses if deviation keeps widening. Best for range-bound pairs.
MA Crossover with Volume Filter
Enter when a fast MA crosses above a slow MA and volume clears your threshold. Exit on crossback or a trailing stop. Helps avoid weak moves in thin liquidity.
RSI Divergence Hunter
Scan for bullish RSI divergence after a down move, then enter on confirmation. Take profit at a fixed target and exit early if RSI fails to recover. Works well on 1H to 4H charts.
Donchian Channel Turtle
Buy on a Donchian breakout and pyramid carefully as the trend continues. Exit on a lower channel break or a time stop. Add position caps to limit drawdowns.
Volatility Regime Switcher
Trade only when volatility fits your regime. Enter on ATR expansion breakouts and reduce size or pause when ATR compresses. Exit with a time stop if momentum fades.
Mean Reversion with Z-Score
Buy when price falls below a moving average by a set Z-score. Sell on mean reversion and cut losses if deviation keeps widening. Best for range-bound pairs.
MA Crossover with Volume Filter
Enter when a fast MA crosses above a slow MA and volume clears your threshold. Exit on crossback or a trailing stop. Helps avoid weak moves in thin liquidity.
RSI Divergence Hunter
Scan for bullish RSI divergence after a down move, then enter on confirmation. Take profit at a fixed target and exit early if RSI fails to recover. Works well on 1H to 4H charts.
Donchian Channel Turtle
Buy on a Donchian breakout and pyramid carefully as the trend continues. Exit on a lower channel break or a time stop. Add position caps to limit drawdowns.
Stop Loss and Trailing Take Profit
Wrap any entry with a fixed stop and a trailing take profit. Tighten the trail after a profit threshold is reached. Useful for protecting gains during fast reversals.
Kalman Filter Trend
Use a smoothed trend estimate to reduce noise. Buy when price crosses above the filter and exit on a cross below. Pair with a volatility filter to avoid chop.
Grid Trading
Place layered buys and sells inside a defined range. Re-center the grid if price drifts beyond a threshold. Add a kill switch if volatility spikes past your limit.
Breakout + Retest
Enter after a breakout, then wait for a retest to confirm support. Set a stop below the retest low and scale out at two profit targets. A cleaner way to join trends.
Funding Rate Hedge
Monitor funding and hedge when rates become extreme. Offset spot exposure with perps to reduce directional risk. Remove the hedge when funding normalizes.
Unlike DIY stacks that break when APIs change, Coinrule centralizes execution and monitoring. You can run the same rule on OKX and KuCoin, or isolate experiments to one venue for cleaner comparisons. Set per-exchange limits, like max 3 open positions on one account and max 6 across all accounts. Logs and notifications make it easy to review what happened and why.
Good automation is not only about entries. Add simple, interpretable filters that often improve real-world performance: trade only during high-liquidity hours, avoid weekends, or require spread to stay below 0.15%. You can also gate trades by market structure, such as only buying after a higher high forms on 4H candles. These constraints reduce false positives without needing complex models in production.
Instead of waiting weeks to ship code, Machine learning trading works best when you shorten the loop from idea to live test. Build a rule, backtest it, then paper trade or go live with small size and strict limits. Start with 0.5% per trade, cap daily loss at 1.5%, and review results weekly instead of reacting to every tick. Over time, you can iterate parameters, add cooldowns, and improve stability without rewriting code.
Your funds stay where you trade. Coinrule connects to centralized exchanges via API keys with configurable permissions, and you can disable keys anytime. For on-chain automation, smart accounts and session keys help limit what an automated session can do. This structure keeps execution flexible while maintaining strong control over access.
Launch BotFAQ
Sofia G.
Jason C.
Lars H.
Priya K.
Diego R.
Yuki T.
Treat each rule like an experiment with a hypothesis and a metric. Track win rate, average win to loss, maximum drawdown, and time in market. Keep one change per iteration so you know what improved results. This approach makes your automation portfolio easier to manage as it grows.
If a strategy only works with perfect parameters, it is fragile. Use wider bands, fewer indicators, and realistic stops so performance does not collapse when the regime shifts. A good test is to vary thresholds by 10% and see if results remain similar. Robustness matters more than a perfect backtest curve.
A model can tell you what to do, but execution decides how well it performs. Add limit orders when spreads widen, or use market orders only when liquidity is strong. Schedule rechecks every 15 minutes instead of reacting to every tick. These choices often improve real outcomes more than adding features to a model.
Single-trade stops are not enough when multiple bots run at once. Set total exposure caps, correlation-aware limits, and a global kill switch for extreme volatility. For example, pause all new entries if BTC drops 7% in 1 hour, then resume after a 2-hour cooldown. Centralized controls keep the whole system coherent.
Start with a proven template, then adapt it to your signal and timeframe. Swap the entry trigger, keep the exits, and test. Once it behaves well, clone it for different coins with adjusted sizing. Over time you will build a library of strategies you actually understand.
Connect an exchange, choose a template, and add your signal filter in minutes. Then backtest, start small, and scale only after consistent results. Coinrule is built for disciplined automation with clear controls and transparent execution. When you are ready, you can expand to more rules, more venues, and more advanced risk logic.
Get Started
