Real-time slippage predictions for crypto, powered by machine learning trained on live order book data. In 2024, traders lost $2.7 billion to slippage.
| Size | Book CostBook | Total (bps) | |
|---|---|---|---|
| $100 | 0.0812 | 1.9026 | |
| $1K | 0.4231 | 2.2445 | |
| $10K | 2.1847 | 4.0061 | |
| $100K | 8.7320 | 10.5534 | |
| $1M | 31.4500 | 33.2714 |
Showing 10 of 20 trade sizes
The Pipeline
We stream depth snapshots, aggregated trades, and best bid/ask from Binance WebSockets. Every tick is captured.
100ms tick rate · 3 streams per symbol
A rolling 60-second window feeds into our feature pipeline: order book imbalance, trade flow pressure, volatility regimes, and cross-level depth ratios.
57 features · 1s recompute cycle
An AI model predicts the execution cost for each of 20 trade sizes, from a $10 retail order to a $1M institutional block. Results stream to you over WebSocket.
20 trade sizes · buy + sell · every second
1s
prediction refresh
57
ML features
$10–$1M
trade sizes
Features
Compare bid and ask slippage simultaneously. Spot market asymmetry at a glance — thin asks with deep bids mean different costs for buyers and sellers.
Buy
$100 0.0812 bps
$10K 2.1847 bps
Sell
$100 0.0734 bps
$10K 1.9201 bps
From retail to institutional. See exactly where the order book gets thin.
$10
$18
$34
$62
$113
...
$2.3K
$4.3K
...
$48K
$89K
$163K
$298K
$546K
$1M
Predictions stream over WebSocket in real time. No page refreshes, no polling. The data is always live.
<1s
Enter any dollar amount and get an interpolated slippage estimate for both buy and sell sides instantly.