#microstructure

The overnight-gap behaviour day-traders actually trade, computed live from Yahoo Finance daily OHLC — no key, nothing stored. A gap is the jump between yesterday's close and today's open — the move that happens while the market is shut, on overnight news and futures drift. Traders live and die on two questions: how often does a name gap, and does the gap fill (price retraces to yesterday's close) or run (it keeps going). This API answers both with hard frequencies. For each instrument it returns how often it gaps up and down beyond a configurable threshold, the average size of up- and down-gaps, the gap-fill rate (the share of gaps where price traded back through the prior close intraday — for an up-gap, the day's low reaching the prior close), and the continuation rate (how often the day closes in the direction of the gap rather than fading it), plus the largest recent gaps. The asset endpoint returns one instrument's full gap profile with its biggest recent gaps; the screener endpoint ranks a universe of liquid stocks and ETFs by gappiness or gap-fill rate, surfacing the names that gap most and the ones whose gaps reliably fill. This is the opening-gap / overnight-jump microstructure cut — distinct from the price, candlestick-pattern, volatility and risk APIs in the catalogue. It is what happens between the close and the open.

api.oanor.com/gapstats-api

Who is actually trading a pair — retail or whales — read from the composition of Binance's aggregated trade tape by trade size, no key, nothing stored. Order flow tells you the net direction; this tells you the size profile behind it: whether a move is driven by a swarm of small retail prints or a handful of large institutional ones, often the more important signal. The distribution endpoint scans the recent aggregated trades for a pair and buckets them into size cohorts (micro under $1k, retail $1k-$10k, mid $10k-$100k, whale over $100k), returning each cohort's trade count, volume in base and quote and its share of total volume, plus the whale-volume share — the single read on how institutional the flow is. The percentiles endpoint returns the trade-size percentiles (p50, p90, p99) and the average, median and largest trade. The symbols endpoint lists tradable pairs. This is the trade-size composition / participant-mix analytics cut for crypto — distinct from the order-flow / CVD API (which measures buy-versus-sell direction), the order-book depth, the slippage and the price APIs in the catalogue. Pairs are Binance symbols (BTCUSDT) or a coin=BTC&quote=USDT form.

api.oanor.com/tradesize-api

Who is actually hitting the market — buyers or sellers — read live from Binance's aggregated trade tape, no key, nothing stored. Every trade carries a flag for which side was the aggressor: a taker buy lifts the ask, a taker sell hits the bid. Summing those over a window gives order flow — the net buying or selling pressure that price action follows — and its running total is the Cumulative Volume Delta (CVD), the metric order-flow traders watch to spot absorption and divergence. The flow endpoint scans the recent aggregated trades for a pair (up to 5,000) and returns the taker-buy and taker-sell volume in base and quote, the delta (buy minus sell), the CVD over the window, the buy/sell ratio, the share of volume that was buying, a net-pressure label and the time span covered. The large endpoint surfaces the big prints — single aggressive trades above a notional threshold — and tags each as a taker buy or sell, so you see the whale orders moving the tape, with the buy- and sell-side large-trade totals. The symbols endpoint lists tradable pairs. This is the trade-flow / CVD microstructure analytics cut for crypto — distinct from the raw recent-trades feed, the order-book depth and the price, ticker and slippage APIs in the catalogue. Pairs are Binance symbols (BTCUSDT) or a coin=BTC&quote=USDT form.

api.oanor.com/orderflow-api