A friend of mine named Alex runs a small prop desk in Ho Chi Minh City. He texted me two months ago: "Bro, I back-tested the Bitsgap COMBO bot for a week and it's printing money. 40% monthly. Should I put $5k in?"
I told him to wait. That I'd run it myself first. Smaller size, real money, no shortcuts.
Here's what I found.
What the COMBO bot actually does
Bitsgap calls it a "COMBO" because it stacks two strategies on top of each other: a grid layer that catches oscillations, and a trend-following layer that adjusts the grid's midpoint as price moves. On paper, it handles both ranging and trending markets, which is the single biggest problem with a regular grid bot.
Regular grid bots die in trends. The price runs out of your range and the bot stops trading. The COMBO tries to solve that by shifting its position as the trend develops. You can run it on spot or on futures. Most people run it on futures because leverage amplifies the grid's small per-trade returns into something that looks meaningful.
The COMBO bot is available on Bitsgap's Advanced plan at $55/month (billed annually), or $69/month if you pay monthly. The Basic plan at $23/month doesn't include futures bots at all. So if you're here because of the COMBO, you're already looking at at least $55/month out the gate.
There's a 7-day free trial with no credit card required, which is how I started.
My test: 45 days, $300, 5x leverage on ETH/USDT
I set the bot up on Binance Futures in mid-April. ETH was trading around $1,820. I allocated $300 USDT, set leverage to 5x (giving me $1,500 in position exposure), and let Bitsgap's AI Assistant suggest the grid parameters. It recommended 28 grid levels across a range of roughly $1,650 to $2,050.
First 10 days: ETH stayed mostly between $1,780 and $1,920. The bot was active, placed 41 trades, collected small profits on both sides. Running P&L was around +$18. Looked fine.
Day 14: ETH dropped hard to $1,620. Below the bot's lower grid boundary. Trading stopped. The bot was holding a full long position at the bottom of the range, worth $1,500 notional, against my $300 margin. Unrealized loss: roughly -$78 at the trough.
I didn't close it. I waited.
ETH recovered to $1,780 by day 23. The bot resumed trading. By day 45 I closed everything. Final result: +$11.40 net after all fees, which is about 3.8% on my $300 margin. Not the 40% monthly Alex saw in his backtest. Not even close.
The demo gap is real
Here's the part that actually matters. Before putting in real money, I ran the same settings in Bitsgap's demo mode for one week. The demo showed +$43.20 over 7 days. Annualized, that's the kind of number that makes people text their friends.
The demo doesn't simulate slippage, execution latency, or real-time funding fees. It assumes every order fills at exactly the limit price. In live trading, especially during the ETH drawdown I described, orders around the range boundaries were partially filled or missed entirely as price whipped through them. The demo showed 41 clean fills in 10 days. Live got 34.
This is not a Bitsgap-specific problem. Every backtesting and demo system has this gap. But Bitsgap's demo results look particularly good because the COMBO bot's design naturally generates a lot of small wins in the demo's idealized fill model. Real markets are messier.
Demo mode is for learning the controls, not for predicting returns. Run it for two weeks to understand the bot behavior, then set your real expectations at roughly 30-50% of what demo showed.
What the platform gets right
I don't want to sound too harsh. Bitsgap's execution interface is genuinely well-designed. Setting up the COMBO took about 12 minutes on my first attempt, including reading the help docs. The AI Assistant for parameter suggestions is actually useful as a starting point (just don't treat its suggestions as optimal).
The backtesting feature is solid. You can test up to 180 days of history on the Advanced plan. I ran ETH back to January and saw exactly the kind of range I was hoping for. The issue is that backtests assume the same clean fills the demo does. But at least 180 days is enough historical depth to spot whether your chosen asset actually ranges at all, which is the first filter before you run any grid-style bot.
Bitsgap connects to 17 exchanges. I tested on Binance but also had it connected to OKX for monitoring. No issues with the API setup. The IP whitelisting step is clearly documented, which is not something you can say about all platforms.
The mobile app exists and is functional. I checked positions from my phone when I was out. No crashes, no strange sync issues. Basic but solid.
Where it struggles
Funding fees are the hidden enemy on any futures bot. During the 45 days I ran the COMBO, ETH's funding rate averaged about 0.022% every 8 hours. That's roughly 2.4% annualized just in funding drag, not a devastating number, but it's constant and invisible if you're not watching. On days when the market was bullish and funding spiked to 0.06-0.08%, the bot was paying more in funding than it was earning from the grid trades.
The platform doesn't surface funding fee drag prominently in the P&L display. You have to go into the fee breakdown yourself to see it. Beginners won't do that.
The bot also doesn't auto-adjust when price breaks out of range. You have to manually intervene, stop the bot, reset the range, and restart it. Some competing platforms have range auto-extension built in. Bitsgap doesn't, at least not on the COMBO bot as of June 2026. (Honestly the manual step is fine if you're checking your positions daily. It's only a problem if you want fully passive operation.)
The $55/month minimum to access futures bots is a real threshold. If you're running $300 in capital like I was, the monthly fee is 18.3% annualized overhead before you make a single trade. You need meaningful capital, probably $1,500 or more, before the math starts to look reasonable. The Bitsgap pricing page makes the plan tiers clear enough, but it doesn't tell you this.
Who should actually use the COMBO bot
If you've been trading on Bitsgap for a while and want to add leverage to your grid strategy, the COMBO is a decent tool. The interface is clean, the setup is fast, and the exchange connectivity is reliable.
If you're new to bots, start with spot grid on a well-known pair. Get comfortable with what ranging behavior looks like, what happens when price breaks a range, and how long recovery takes. Then think about futures.
Don't run the COMBO on a volatile alt with 10x leverage because a backtest looked good. That's the setup for a blown account. ETH at 5x is already spicy when it drops 8% in a day. I've seen people on forums describe losing 60% of their margin on a single ETH wick with this bot running. I didn't verify those accounts, but they're plausible.
For a comparison against a platform with deeper DCA bot features and lower entry cost, see our 3Commas vs Bitsgap breakdown. If you're not sure which bot type fits your style, the bot match quiz takes about five minutes and gives you a specific recommendation.
[AFFILIATE: Bitsgap]
Alex did not put $5k in. He ran the demo for another two weeks, saw the same 40% number, and then asked me why my live results were so different. We spent an afternoon going through the fee logs. He's now running the COMBO with $800 at 3x leverage on BTC, which is the most boring version of the setup. It's been three weeks and he's up $19. That's realistic.
