Buildix API for Crypto Orderflow: Complete Developer Reference Guide (2026)

The crypto orderflow data layer is not a commodity. Most public APIs give you OHLCV candles and basic ticker data. A handful give you funding rates and open interest. Almost none give you Volume-synchronized Probability of Informed Trading, multi-level Order Flow Imbalance, cross-exchange Cumulative Volume Delta confirmation, or predictive liquidation maps. Those are the data layers that institutional desks pay six figures a year for through Bloomberg, Refinitiv, or direct exchange data feeds. The Buildix API exposes them at a fraction of the cost, on a 24/7 REST and WebSocket interface, across 530+ perpetual futures pairs.

This is the complete developer reference for the Buildix API as of May 2026. If you are building a trading bot, an automated screener, a custom dashboard, a research backtest pipeline, a Telegram alert relay, or a quantitative strategy that needs orderflow inputs, this guide tells you what is available and how to call it.

Authentication and access. The API is gated by tier. The Free tier exposes a small set of read endpoints with rate limits suitable for casual exploration. The Trader tier opens up most market data endpoints. The Pro tier unlocks signal scores and historical backfills. The Whale tier ($79 a month) unlocks the full surface, including the predictive liquidation map, smart money positioning, real-time WebSocket streams, and write endpoints for alert and watchlist management. Every request requires an API key passed in the Authorization header. Generate yours from your account settings page once you log in.

Base URL and rate limits. The base URL is https://api.buildix.trade. All responses are JSON. All timestamps are ISO 8601 UTC unless otherwise noted. Rate limits are tier-dependent and enforced on a sliding window basis. Free tier accounts get 60 requests per minute. Trader gets 300. Pro gets 1,200. Whale gets 6,000 plus dedicated WebSocket connections. Standard rate-limit headers (X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset) are returned on every response, so your client can adapt without hitting hard 429 errors.

Below are the categories of data the API exposes, with a representative endpoint and example for each.

Pair metadata and screener data. The pair endpoint returns the canonical metadata for every pair we track, including the exchange, the contract type (perp, spot, HIP-3 builder), the minimum order size, the maximum leverage, and the current tick size. The screener endpoint returns the same data plus the live computed metrics for each pair, which is the same dataset that powers the public Buildix screener. A typical curl request looks like this. curl -H "Authorization: Bearer YOUR_API_KEY" "https://api.buildix.trade/v1/screener?exchange=hyperliquid&limit=100". The response is a sorted list of pair objects, each carrying the latest score, CVD, OBI, VPIN, volume, and funding rate. In Python with requests, the same call is two lines. import requests; r = requests.get("https://api.buildix.trade/v1/screener", headers={"Authorization": "Bearer YOUR_API_KEY"}, params={"exchange": "hyperliquid", "limit": 100}); data = r.json(). The structure is documented inline in the response under a self-describing schema field, which means your code can adapt to schema changes without a rebuild.

VPIN and CVD time series. VPIN and CVD are returned as time series with configurable bucket size and lookback. The default VPIN bucket is one minute, computed using the standard Easley-Lopez de Prado clock-time bulk volume classification. The CVD endpoint returns both raw and bucketed values. Time series queries support exchange filtering, which is how cross-exchange CVD confirmation is computed downstream. A Pro-tier developer can fetch BTC-USDT VPIN on Binance and Bybit simultaneously and compute the divergence client-side, or they can call the cross-exchange endpoint directly to get the precomputed score. We have a dedicated tutorial on fetching VPIN and CVD data via Python that walks through this pattern with full code, and a similar walkthrough for fetching real-time data over WebSocket.

Order book and depth data. The orderbook endpoint returns the current snapshot at configurable depth (5, 10, 20 levels). The OBI endpoint returns the precomputed Order Book Imbalance ratio at each of the three depth levels we compute simultaneously, which is what powers the multi-level OBI input to the V5 signal engine. The OFI endpoint returns the Order Flow Imbalance time series, computed as the change in net resting depth at each price level over a rolling window. For developers building latency-sensitive strategies, the WebSocket equivalent of these endpoints streams updates at the same cadence as our internal pipeline, which is sub-second on Hyperliquid and most centralized venues.

Funding rates and open interest. The funding endpoint returns the current funding rate for any pair on any supported venue, plus the trailing 7-day, 30-day, and 90-day distributions. Z-scores against trailing distribution are precomputed, which means your bot does not need to maintain its own statistical baseline. The OI endpoint returns open interest in both base currency and notional USD, plus the change rate over rolling windows. Cross-venue aggregation is available, which is useful for funding rate arbitrage strategies where the spread between venues is the alpha.

Liquidation map. The liquidation map endpoint is one of the most popular Whale-tier features. It returns the predictive liquidation cluster data, computed by aggregating positions across the available on-chain and exchange data sources, then projecting where forced liquidations would trigger if price moves through specific thresholds. The output is a heatmap-ready array of price-level buckets with notional liquidation size in each. We covered the predictive liquidation map API in detail in a dedicated post for developers building liquidation-aware bots.

Smart money and whale tracking. The smart money endpoint returns the live long-short ratio of the top wallets we track on Hyperliquid, with rolling windows from 1 hour to 7 days. This is the same data behind the smart money positioning API on Hyperliquid that we documented separately for whale-following automation. The whale endpoint streams large trades as they print, classified by size percentile against the trailing 24-hour median. Both endpoints are WebSocket-streamed for real-time consumers, with REST endpoints for batch backfills.

Signal Engine V5 outputs. The signals endpoint returns the live composite V5 score for every pair, plus the breakdown by component (CVD, VPIN, OBI, OFI, whale flow, funding pressure, liquidation squeeze, volume context multiplier). This is the same engine documented in the Signal Engine V5 deep dive. Pro-tier users can subscribe via WebSocket to receive score updates the moment they cross thresholds. Whale-tier users can configure custom score formulas client-side and receive alerts on those derivatives.

Alert and watchlist management. Whale-tier accounts can manage their alert configuration via the API. POST to /v1/alerts/subscribe to add a pair to the alert list with a custom threshold and direction. POST to /v1/watchlists to create or modify a custom watchlist. These endpoints are how you script multi-account configurations or sync watchlists across multiple devices. All user data is stored against the account in the cloud, which means an alert configured via API is identical to one configured via the web UI.

Historical backfills. Pro-tier and Whale-tier accounts get historical data access. The /v1/historical endpoint accepts a pair, a metric, a start date, and an end date, and returns the aggregated time series. Granularity is configurable from 1-minute up to daily. We retain orderflow component history for 90 days for Pro and 365 days for Whale, which covers most backtesting use cases. Beyond 365 days, the data is available on request as a one-time CSV export for an additional fee.

WebSocket streams. WebSocket connections are available at wss://stream.buildix.trade. The connection accepts a subscription message with the streams of interest and returns updates as they happen. Available streams include the screener (full live updates for all pairs), the signal engine (V5 score updates), whale alerts (large prints with classification), liquidation events (forced closures as they occur on supported venues), and pair-specific orderflow (CVD, OBI, OFI ticks for a specific pair). Connection limits are tier-dependent. Pro accounts get up to 5 concurrent connections. Whale accounts get up to 50.

A simple end-to-end example in Python. The following script connects to the WebSocket, subscribes to BTC-USDT signal updates on Binance, and prints the live V5 composite score every time it updates. import asyncio, json, websockets; async def listen(): async with websockets.connect("wss://stream.buildix.trade") as ws: await ws.send(json.dumps({"action": "subscribe", "auth": "YOUR_API_KEY", "streams": ["signals.binance.BTC-USDT"]})); async for msg in ws: print(json.loads(msg)); asyncio.run(listen()). The script runs in 12 lines and produces a continuous stream of orderflow scores. Replacing the streams parameter with a different pair or a different exchange gives you the equivalent feed for whatever you are tracking.

A note on JavaScript and other languages. The API is HTTP and WebSocket only. No SDK is currently published, but the contract is small enough that a thin wrapper in any language is straightforward. We recommend axios or native fetch in Node.js for REST, and the standard ws package for WebSocket. The API responses include schema metadata, which means you can use code generation against a single sample response to produce typed clients in TypeScript, Go, Rust, or whatever your stack prefers.

Why developers pay for this. The Buildix API is not the cheapest way to get crypto OHLCV data. Anyone can pull that for free from the major exchanges directly. The reason developers pay is the orderflow layer. Computing VPIN correctly across 530+ pairs in real time is a non-trivial engineering problem. Maintaining cross-exchange CVD synchronization, classifying whale trades reliably, projecting liquidation cascades, and keeping the data fresh at sub-second cadence is multiple full-time engineering hours per week. The API exists so that a developer building a trading product or a quantitative strategy can pay $79 a month for the data layer instead of building it from scratch.

The integrated AI Strategy Advisor with BYOK across OpenAI, Anthropic, Google, Groq, Mistral, and Ollama is also available via API for developers who want to run model-driven analyses on top of the orderflow data without paying separate model API fees. You bring your own key, and Buildix routes the request without taking a margin on the model usage.

Hyperliquid is supported from day one, which means HIP-3 builder pairs (S&P 500, oil, gold, FX, NVDA, TSLA) are available through the same endpoints as the spot and standard perp markets. This still distinguishes Buildix from most analytics APIs, where Hyperliquid is either missing or added only after months of community pressure.

To get started, sign up at buildix.trade, generate your API key from the account settings page, and try the screener endpoint first. From there, build whatever you want. The data is the layer that institutions guard most carefully and retail almost never reaches. With Buildix, retail reaches it for the price of a streaming service.