Uncle Sol's arb desk logged 59 cross-exchange spread events on August 3, 2026, topped by a 21.28% BLESS gap between Bybit and Bitget and a cluster of four ICNT spreads spanning Coinbase, Bybit, KuCoin, Binance Futures, and Gate Futures.
Fifty-nine arbitrage events crossed the scanner today, and the top of the board belonged to BLESS, which tore open a 21.28% spread between Bybit ($0.017931 bid-side buy) and Bitget ($0.021747 sell-side print). That is not a typo-tier spread you see on BTC or ETH pairs — this is micro-cap territory, where thin books and fragmented price discovery across venues routinely blow past what's possible on majors. Right behind it, ICNT put on a clinic in cross-venue mispricing, showing up four separate times in the top ten with spreads ranging from 9.22% up to 21.01%, bouncing between Coinbase, Bybit Spot, Bybit, KuCoin, Binance Futures, and Gate Futures over the course of the session.
For the arb desk, today's tape is a reminder that the best edges live where liquidity is worst. Every name on this list — BLESS, ICNT, BEAT, AIO, UAI, 1000RATS — trades in the sub-$4 range with several priced in fractions of a cent, and 1000RATS clearing at $0.000058 on Binance Futures. That combination of low float and multi-venue listing fragmentation is exactly the setup that produces 8-21% headline spreads. The catch, as always, is whether size can actually clear the book at the quoted print. Below we break down the top five, the exchange-pair patterns behind them, and the fee math that separates a real trade from a phantom spread.
Bybit is the common thread through today's board — it appears in six of the top ten events, almost always as the buy-side (cheap) leg against Bitget, KuCoin, Binance Futures, and Gate Futures, with the single exception of the Coinbase-ICNT pair where Bybit was the expensive sell side. That pattern suggests Bybit's spot desk is either slower to reprice small-float alts intraday or simply carries less aggressive market-making depth on these specific tickers than its larger-volume rivals — worth flagging as a recurring source of edge, not a one-off.
A second, tighter pattern shows up in Bitget vs. Bitunix: AIO (8.52%, buy Bitget $0.040437 / sell Bitunix $0.042780) and UAI (8.01%, buy Bitget $0.501040 / sell Bitunix $0.541155) both fired the identical direction — Bitget cheap, Bitunix rich. Two independent tickers running the same buy/sell venue pairing in the same session is a stronger signal than either spread alone; it points to Bitunix consistently marking these smaller alts a touch hot relative to Bitget's book, possibly due to lighter market-maker coverage on Bitunix for this tier of token. Elsewhere, BLESS showed the same Gate Futures-cheap-vs-Binance Futures-rich pattern (8.25%) that echoed its own Bybit-vs-Bitget spot spread, and 1000RATS ran Binance Futures cheap against Bitget rich (7.78%) — Binance Futures showed up on both the buy and sell side across different tickers today, underscoring that it's less about one venue being universally mispriced and more about liquidity depth mismatches on a per-token basis.
Every name on today's board is a micro-cap, and that changes the speed-vs-size calculus completely versus arbing BTC or ETH. On majors, a 0.1% spread with $10M of depth is a size game — you scale up until fees eat the edge. Here, the headline spreads are 8-21%, but the depth behind them is unknown (today's volume feed returned $0.0M across the board for pump/dump and buy/sell pressure, meaning no confirmed flow data backs these prints). That flips the priority: speed and small clips first, size only after you've confirmed the book can actually absorb you.
Practical approach: start with a probe order sized to roughly 1-2% of what you think the spread is worth capturing, execute both legs, and only then check your realized fill price against the quoted spread. If slippage on that probe is under 1-2%, scale up in tranches rather than firing one large order — market orders on thin books like BLESS ($0.0179 handle), 1000RATS ($0.000058 handle), or AIO ($0.040 handle) will walk the book fast, and by the third or fourth tranche you may be paying materially worse than the first print. For the ICNT cross-market pairs (spot vs. futures), keep position sizing tighter still, since you're also carrying basis and margin risk on top of execution risk. As a rule of thumb on this tier of token: treat anything past $2,000-$5,000 notional per leg as size that needs a depth check first, not a market order.
Walking through the top trade — BLESS, buy Bybit $0.017931, sell Bitget $0.021747, gross spread 21.28% — on a $5,000 notional clip: at $0.017931 you pick up roughly 278,900 BLESS. Apply a 0.10% Bybit spot taker fee (~$5) on entry. Selling that same 278,900 BLESS on Bitget at $0.021747 grosses roughly $6,065; apply Bitget's 0.10% taker fee (~$6) on exit. Add a rough flat withdrawal cost if you're transferring the asset between venues rather than trading from pre-funded balances — call it $1-3 depending on network conditions at the time. Net profit lands around $1,045-$1,053 on the $5,000 clip, or roughly a 20.9-21.1% net return — fees and a transfer cost only shave about 0.2-0.4 percentage points off the 21.28% headline here because the spread is so wide relative to typical CEX taker fees (0.05-0.10% on futures, 0.10-0.60% on spot depending on tier and venue).
Today's volume tracker came back flat across the board — $0.0M in pump volume, dump volume, buy pressure, and sell pressure — meaning none of these 59 events carry confirmed on-exchange flow data behind them. Treat every spread on this list as indicative until you've pulled live order book depth yourself; a headline percentage with no volume behind it is a starting point for research, not a green light to deploy capital.
ICNT is the name to have on the watchlist first thing tomorrow — four separate spread events across five different venues (Coinbase, Bybit Spot, Bybit, KuCoin, Binance Futures, Gate Futures) in a single session says this token hasn't found its equilibrium price yet, and that kind of dislocation often persists for a day or two before market makers close the gap. Watch Bybit vs. KuCoin and Bybit vs. Coinbase specifically, since both produced double-digit spreads today. BLESS is the second name worth tracking — the Bybit-Bitget and Gate Futures-Binance Futures pairing both fired today, and a token that prints two independent double-digit spreads across four venues in one session is worth a second look at the next session open. Keep Bitget-vs-Bitunix on the radar too, given AIO and UAI both ran the identical pattern — if a third small-cap ticker shows the same Bitget-cheap, Bitunix-rich setup tomorrow, that's a venue-level pattern worth building a standing watch around, not a coincidence.
Best windows to watch: session opens, particularly the Asia open and the US morning, when overnight low-volume drift gets repriced hard against the more liquid venues. Exchange pairs to prioritize for the next session: Bybit vs. Bitget, Bybit vs. KuCoin, Bitget vs. Bitunix, and Binance Futures vs. Bitget — all four produced confirmed spreads today and are the most likely to repeat.
Fifty-nine events, one 21%-plus spread, and a small-cap corner of the market that clearly hasn't settled on a single fair price across venues yet. Size small, check the book before you check the headline number, and don't let a wide spread on a thin coin talk you into a fat order. Watch ICNT and BLESS first thing tomorrow. Arbitrage Hunter — August 3, 2026.