"The most profitable trading bots on Polymarket": two make money trading, two make it from rebates
A viral X article showed four crypto-candle bots with profits from $231K to $1.36M. Those figures leave out fees. After fees, the two biggest bots barely break even or lose money, and Polymarket's rebates are what keep them ahead.
The claimFour bots are among "the most profitable trading systems on Polymarket", in "the 6 and 7-figure range", and they "continue to generate positive PnL". — X (@RetroValix)
The short version
The article, "The Most Profitable Trading Bots on Polymarket Right Now", was posted on X by @RetroValix on August 24, 2026 and has been viewed about 2.6 million times. The author says they used Claude to analyse every trade of four bots that trade the crypto "Up or Down" markets, from 5-minute to 4-hour candles, and describes each one's strategy. Its evidence is a set of profile screenshots: Bonereaper $1,362,556, 0xb55… $965,389, mo-money $301,377 and almach $231,157 all-time profit. It says these systems "continue to generate positive PnL" where competitors have stopped. The article sells nothing; it ends by asking readers to follow the author.
Polymarket publishes a daily breakdown of every wallet's P/L: trading before fees, fees paid, fees refunded, and rebates and rewards. On August 23–24, mo-money's trading before fees stood at $301,766 and almach's at $231,292, the screenshot figures to within a day. The profile number is profit before fees. Bonereaper's and 0xb55's screenshot figures are higher still than that measure on August 24 ($1.16M and $949K), and we could not find what they count. For those two, fees are the whole story. mo-money and almach mostly rest their orders, paying taker fees on 17–19% of the money they spend buying, and have paid $57K and $48K in fees in total. Bonereaper (56% taker since May) and 0xb55 (76%) have paid $1.44M and $1.31M; Bonereaper's figure is after Polymarket refunded it another $907K in the spring.
After fees, Polymarket's own leaderboard puts the four at −$134K (Bonereaper), +$28K (0xb55), +$458K (mo-money) and +$369K (almach) all time. All four accounts are still ahead overall, because Polymarket pays rebates on these markets: $613K to Bonereaper and $544K to 0xb55, against $66K and $63K to the other two. Counting everything, the four are up $479K, $573K, $525K and $432K. These are real, profitable accounts. But for Bonereaper and 0xb55, the profit is Polymarket's rebates, not their trading.
The record since the article separates them further. From August 24 to October 27, mo-money's trading made $181K after fees and almach's $158K. Bonereaper's trading lost $110K and 0xb55's lost $89K, even as the before-fee profile figures the article quoted kept climbing, by $149K and $385K. Rebates of $197K and $257K kept both accounts up. Bonereaper's trading after fees has lost money in every month since May; 0xb55's has in September and October. Rebates are paid on trading activity, and these two bought $157 million and $69 million of contracts. A smaller copy of either strategy would pay the same fees without collecting rebates on that scale.
Some of the article's descriptions hold up in the 30 days before it was posted: 0xb55's median gaps between first buys on each side (51 seconds on 5-minute markets and two and a half minutes on 15-minute ones, against the article's 52 seconds and about three minutes), and almach's near-equal Up and Down sizes on hourly markets. Others don't: Bonereaper bought both sides in 82% of its 5-minute markets, not 55%, at a combined price close to $1.00, not $1.08. The descriptions also don't show where the money came from. almach, described as building Up+Down pairs gradually, made its pre-article profit in markets where it bought only one side (+$244K), while markets where it held both sides lost $27K. 0xb55's matched pairs, the "complete sets" the article credits for its edge, made $646K before the article and lost $547K after. Verdict: misleading. The bots are real and all four accounts are up, but the headline figures are profit before fees, and for the two largest, trading after fees is a loss covered by Polymarket's rebates.
Technical appendix
Exhibit A Profit as shown, and after fees
Polymarket's own figures for each wallet (data-api /v2/user-pnl and /v2/leaderboard, read October 27, 2026). "Before fees" is the snapshot at 00:00 UTC on August 24, about 18 hours before the article was posted.
| Shown in the article | Trading before fees, Aug 24 | Fees paid, all time | Trading after fees, all time | Rebates and rewards | Total | |
|---|---|---|---|---|---|---|
| Bonereaper | $1,362,556 | $1,161,676 | $1,444,183* | −$133,985 | $613,209 | $479,224 |
| 0xb55… | $965,389 | $949,292 | $1,305,620 | +$28,748 | $543,847 | $572,596 |
| mo-money | $301,377 | $301,766 | $57,319 | +$458,083 | $66,465 | $524,548 |
| almach | $231,157 | $231,292 | $48,438 | +$369,192 | $62,709 | $431,901 |
* After $906,915 that Polymarket refunded to Bonereaper in March and April; fees charged before refunds were $2,351,098. "Rebates and rewards" are maker rebates, taker rebates, liquidity rewards and referral income. The leaderboard's all-time figures (−$134,104; +$27,572; +$457,808; +$368,993) are the after-fee column without the few dollars of open positions.
Exhibit B Before and after the article
Polymarket's own figures, split at the start of August 24 (the article went up at 18:10 UTC that day) and read on October 27.
| Trading after fees, up to Aug 24 | Trading after fees, since | Rebates and rewards, since | Total, since | Profile figure (before fees), since | |
|---|---|---|---|---|---|
| Bonereaper | −$23,977 | −$110,008 | +$197,247 | +$87,239 | +$148,522 |
| 0xb55… | +$117,724 | −$88,976 | +$256,743 | +$167,768 | +$385,076 |
| mo-money | +$277,371 | +$180,712 | +$32,554 | +$213,266 | +$213,636 |
| almach | +$211,340 | +$157,852 | +$30,190 | +$188,042 | +$186,338 |
Our own count of Bonereaper's cash in and out over the same period agrees: −$109,829 from trading.
Exhibit C Month by month: trading after fees
Change in each wallet's after-fee trading P/L per calendar month (UTC), from Polymarket's daily series. Rebates are not included. October runs to the 27th.
| Month | Bonereaper | 0xb55… | mo-money | almach |
|---|---|---|---|---|
| March (from the 25th) | +$72,897 | — | — | — |
| April | −$101,060 | — | — | — |
| May | +$116,518 | +$49,490 | — | −$420 |
| June | −$54,306 | −$10,423 | +$32,799 | +$27,350 |
| July | −$59,538 | +$43,601 | +$141,652 | +$95,495 |
| August | −$65,051 | +$35,852 | +$130,413 | +$106,544 |
| September | −$33,819 | −$39,693 | +$94,152 | +$92,350 |
| October | −$9,624 | −$50,079 | +$59,056 | +$47,873 |
Bonereaper's after-fee trading peaked at +$101,752 on June 19. mo-money placed a few sports bets in February and began trading crypto candles in June.
Exhibit D The article's behaviour figures, re-measured
The article does not say which period it measured. We used the 30 days before it was posted (July 25 – August 24, by when each market's trading began). "Both sides" counts markets where the bot bought Up and Down, out of markets where it bought anything. Combined price is the median of (average price paid for Up + average price paid for Down) per two-sided market, before fees.
| Bot | Figure | Article | Our count |
|---|---|---|---|
| Bonereaper | 5m markets with both sides | ~55% | 82% |
| 5m gap between first buys on each side | ~26 s | 21 s | |
| 5m smaller side ÷ larger side | ~45% | 53% | |
| 5m combined price (average) | ~$1.08 | $1.01 | |
| 0xb55… | All markets with both sides | ~75% | 84% |
| 15m / 1h / 4h with both sides | 79% / 87% / 100% | 90% / 96% / 95% | |
| Gap between first buys, 5m / 15m | ~52 s / ~3 min | 51 s / 2.5 min | |
| Combined price, 1h / 4h | ~$0.93 / ~$0.91 | $0.96 / $0.97 | |
| mo-money | 5m / 1h / 4h with both sides | 32% / 67% / 85% | 39% / 73% / 85% |
| 5m smaller side ÷ larger side | ~43% | 67% | |
| 5m combined price (average) | ~$1.07 | $0.97 | |
| 1h smaller ÷ larger / combined price | "almost identical" / ~$0.91 | 93% / $0.95 | |
| almach | 5m / 1h / 4h with both sides | 23% / 72% / 96% | 36% / 70% / 87% |
| 1h smaller ÷ larger / combined price | ~96% / ~$0.97 | 95% / $0.97 | |
| 4h smaller ÷ larger / combined price | ~92% / ~$0.99 | 92% / $1.00 |
Over each bot's whole history before the article, the shares of two-sided markets move by a few points; the hourly combined prices for mo-money and almach come out higher, at about $1.00.
Exhibit E Where the money came from
Our count from each wallet's public activity, market by market: cash in minus cash out, after fees, before rebates. A market counts as before or after the article by when the bot's trading in it began. "Matched pairs" is the part of two-sided positions where Up and Down shares pair off into a set that pays exactly $1 whatever happens; it is valued at $1 minus what the pair cost, fees included.
| One side only | Both sides | of which matched pairs | Total | ||
|---|---|---|---|---|---|
| 0xb55… | Before | −$71,379 | +$187,738 | +$646,056 | +$116,359 |
| After | +$62,295 | −$149,656 | −$546,550 | −$87,361 | |
| mo-money | Before | +$255,906 | +$26,137 | −$12,921 | +$282,043 |
| After | +$86,392 | +$93,933 | +$50,018 | +$180,325 | |
| almach | Before | +$243,524 | −$27,065 | −$2,368 | +$216,509 |
| After | +$80,886 | +$76,830 | +$55,064 | +$157,716 |
0xb55's pairs turned from profit to loss as their median cost, fees included, rose from 98.2¢ before the article to 100.0¢ after. Bonereaper is left out of this table: in March and April its fees do not appear in its trade records, so a per-market count is not reliable for that period. Its pairs cost more than $1 throughout (a median of $1.02 including fees before the article and $1.06 after), consistent with the article's description of a bot that pays for its hedges.
Exhibit F How we counted
We downloaded each wallet's full public activity from Polymarket's data API (/activity, paged by the hour and split further wherever a page filled up, so nothing was truncated): 15.1 million records in all. The wallets: Bonereaper 0xeebde7a0e019a63e6b476eb425505b7b3e6eba30, 0xb55… 0xb55fa1296e6ec55d0ce53d93b9237389f11764d4, mo-money 0x32ed2e546b187ca15e2841edc82b22c713cf8ec3, almach 0x3725d52f3c252e8374999cc8617292ea2608ad88.
- Cross-check. For 0xb55, mo-money and almach, our total of cash in and out plus current positions comes within 1.3% of Polymarket's own total, and our measured taker fees within 1%. For Bonereaper our total is $189K higher than Polymarket's, and all of the gap is in March and April, when Polymarket also refunded it $907K of fees. From May on, the two agree within 5%. Where the two differ, the headline figures above use Polymarket's.
- Taker fees are measured on each buy as the amount paid above price × shares.
- All four bots are still trading. Each holds under $2,500 in open positions, so the totals above are almost entirely settled.