DCA bots

The multiplier is the strategy - everything else is UI

Bot	Type	Price	Trial	DCA	Backtest	Paper	Trade fee	Exchanges	US
Freqtrade	Self-hosted	free	-	yes	yes	yes	-	11	US
Gunbot	Self-hosted	$29/mo	-	yes	yes	yes	-	20	US
Bitsgap	Cloud platform	$23-149/mo	7 days, no card	yes	yes	yes	-	17	US
OctoBot	Self-hosted	$0-29.99/mo	7 days, no card	yes	yes	yes	-	12	US
Coinrule	Cloud platform	0.02% of volume	-	yes	-	-	-	20	US
Superalgos	Self-hosted	free	-	yes	yes	yes	-	15	US
Cryptohopper	Cloud platform	$29-129/mo	3 days, no card	yes	yes	yes	-	16	US
PI Pionex	Exchange-native	0.05% per trade	-	yes	-	-	0.05% per trade	-	US
WunderTrading	Cloud platform	$0-100/mo	7 days, no card	yes	-	yes	-	18	US
Passivbot	Self-hosted	free	-	yes	yes	yes	-	8	US
3Commas	Cloud platform	$15-140/mo	3 days, no card	yes	yes	yes	-	9	US
HaasOnline	Cloud platform	$19.99-249.99/mo	7 days, no card	yes	yes	yes	-	24	US
MI Mizar	Cloud platform	0.1% of volume	-	yes	yes	-	-	10	-
Binance Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
OKX Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
GA Gate.io Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
KuCoin Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
Bitget Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
BY Bybit Trading Bots	Exchange-native	free	-	yes	-	-	-	-	-
TR TradeSanta	Cloud platform	$14-70/mo	14 days, no card	yes	-	-	-	8	US

Live values from our database - details and caveats in each review. "-" = not documented.

A DCA bot is a base order plus a ladder of safety orders that fire as price falls, each sized by a multiplier. Vendors present the multiplier as a profit dial. It is a capital-allocation dial. Our simulation (base $100, 7 safety orders, first deviation 2%, price step scale 1.4x, fills on daily lows - model in research/data/dca_sim.py, no fees, no TP cycles) puts numbers on the ladder:

• Flat 1.0x: $800 covers the same depth the aggressive ladder covers with $25,500. Average entry after 8 fills: -20.1% under base.
• Moderate 1.5x: $4,926 - entry improves to -33.3%. The last two orders alone carry ~35% of capital.
• Aggressive 2.0x: $25,500 - entry reaches -38.6%. The final order alone consumes 50% of everything committed.

That is the trade: 32x capital buys 18.5 percentage points of entry improvement. The community critique has said it plainly since 2022 - safety orders accumulate without closing anything, hence "holding a bag"[6].

There is also an academic floor under plain DCA that the bot marketing skips: Graham described fixed-dollar periodic buying in 1949 as the defensive investor's discipline[1], and Constantinides showed in 1979 that DCA is suboptimal for an expected-value maximizer - it is a risk-policy choice, not a return enhancer[2]. Bots automate that policy and add timed extra fills on top; they do not repeal the underlying result. When a vendor's landing page implies the ladder produces edge, the honest reading is: it produces a different drawdown profile you must pre-fund.

Fees sit on every rung. Each safety buy pays the trading fee[4][5], so a 0.1% venue charges 0.2%+ per average-lowering fill on the round trip that eventually closes. On a deep ladder (8+ fills), fees alone can consume several tenths of a percent of position value - which is why the fee-light native bots and the BNB-discounted tier matter more here than any configurator feature.

Drawdown replays on real BTC data

Across 20 DCA-capable platforms in this lane, the honest question is what your ladder does in real drawdowns. We replayed four historical episodes (position PnL at cycle end, no TP cycles or fees):

Read the columns twice. Higher multipliers improve every percentage - and multiply every absolute number. On 2021-22, the 2.0x ladder lost $14,408 on a $25,500 commitment; the 1.0x ladder lost $533 on $800. A 27x difference in absolute loss, from one settings toggle. Since 2017, BTC has printed 22 drawdown episodes of -20% or worse, including a running -53% since October 2025. Any ladder you configure should be sized against that history, not against the last three months.

Which platform for which case

Zero-subscription DCA with real backtesting: Freqtrade (free, open source, 11 exchanges, dry-run + backtesting[7]) and OctoBot's open-source core are the honest baseline. Superalgos and Passivbot cover scripted and futures-native variants. You trade subscription cost for setup and ops effort - the self-hosted lane details that bill.

Cloud configurators: 3Commas defined the base-order-plus-safety-ladder UX and still prices it at $15/38/105/mo annual ($20/50/140 monthly) with 5/20/1000 bot limits by plan[3]. Bitsgap ($23-149/mo, 7-day trial), WunderTrading (free tier, $0-100/mo, 7-day trial) and Cryptohopper ($29-129/mo, 3-day trial) cover the same pattern with different marketplace angles. Coinrule bills 0.02% of volume instead of a subscription - cheap at low volume, expensive at scale.

No-key and fee-light: Pionex runs its DCA bot inside the exchange at 0.05% per side[4] - no API keys, no subscription; Binance, OKX, Gate and peers ship equivalent native bots. Fewer knobs than the cloud configurators, dramatically lower total cost of ownership.

Tips that actually move the needle

• Compute coverage before deployment. The DCA safety-order calculator turns your deviation/step/multiplier settings into "deepest drawdown covered" - if that number is shallower than BTC's worst historical drawdown, you sized for hope.
• Separate position PnL from capital PnL. Vendors showcase position percentages; your account sees committed capital. A -15.6% position drawdown at 2.0x is -$3,979 on $25,500 - the same absolute hit as -35.2% at 1.0x is -$282 on $800. Different lives.
• Set a max-deviation circuit breaker. Decide the price at which you stop averaging and reassess - otherwise the ladder's logic, not your judgment, owns the last 20% of capital.
• Mind the fee floor. Every safety buy pays the spread and fee[5]; tight deviation steps on 0.1% fees burn more per fill than the averaging gains. Pionex's 0.05%[4] or BNB-discounted Binance (0.075%[5]) change this calculus more than platform choice does.
• Paper trade one drawdown leg, not one uptrend. TP cycles look great in green weeks; the ladder's character only shows in red ones. Freqtrade's dry-run and Bitsgap's paper mode both let you ride a fake -20% week before risking real capital.

Warnings

• Any fixed-return promise in DCA-bot marketing is the oldest red flag in the space - the FTC's line is that only scammers promise fixed returns[7]. Ladders change entry prices, not direction risk.
• The bag is structural. When the ladder empties, the position is at maximum size at the worst average you will ever get. Decide your abandonment rule before the first fill, not during the drawdown[6].
• Running 2025-26 drawdown is live evidence. A -53% episode that started October 2025 is still open as of our last data pull - every ladder built since has been tested by it[8]. Size against it.
• Freemium limits distort risk. Free tiers cap active bots or volume; traders sometimes compensate by overloading each bot with aggressive multipliers. The constraint should limit size, not inflate per-bot risk.
• TP-cycle restarts re-risk everything. Each completed cycle redeploys the full ladder; in a sustained decline, that converts a one-time drawdown into a sequence of them. Consider fixed-size cycles (sell lot, restart smaller) on platforms that allow it.
• Simulated fills use daily extremes. Our model fills on daily lows - optimistic for buys, since intraday paths may not reach your trigger. Real fills land somewhere between our replay and your hopes; both are why coverage margin exists.

What the numbers say about this lane

• 20 bots match the DCA filter on this page; 17 of them are free or carry a free tier, and 7 offer a trial before any commitment.
• Trial lengths run up to 14 days - long enough to paper one drawdown leg if you actually use the window.
• Median entry price across the subscription lanes: ~$21.49/mo/mo - the number your first TP cycles have to beat after fees.
• The DCA bots guide carries the full saturation table and the harmonic-mean argument for why fixed-amount fills average better than fixed-share fills[1].