OHLCV Data - Aggregated Candlestick Data

Pre-aggregated candlestick data (Open, High, Low, Close, Volume) for Hyperliquid markets, exposed through a public Dwellir endpoint over REST, JSON-RPC, and WebSocket.

What is OHLCV Data?

OHLCV data aggregates individual fills into time-based candles that are easier to use for charting, technical analysis, portfolio monitoring, and strategy backtesting than raw execution streams.

This product is built from the same raw fill data Dwellir collects from Hyperliquid and stores canonically, then derives sparse 1s, 1m, and 5m candles from that trade history.

Available Intervals

Larger intervals are intentionally not materialized today. Clients can derive them from the shipped 1s, 1m, or 5m candles.

Available Markets

OHLCV data covers all Hyperliquid market types. Each type uses a different symbol format in the market parameter:

Discovering market symbols

Use Dwellir's Info Endpoint to look up valid symbols before querying candles.

For full details on how to query these endpoints, see Discovering Markets in the candles docs.

If you need to map a market to the underlying Hyperliquid asset index, start from meta, spotMeta, and perpDexs.

Candle Schema

s, i, t, T, o, h, l, c, v, q, n, x

How to Use the Endpoint

Choose the transport that matches your workflow:

• candles for single-market REST lookups, JSON-RPC lookups, and WebSocket live updates
• allCandles for one live stream of changed candles across active markets
• Full-History Exports for ready-made CSV or Parquet files of the full archive for one market and interval
• Archive backfill guide for DIY historical retrieval of one market and one interval
• CSV export guide for DIY flat-file extraction by market and interval
• Parquet export guide for DIY columnar analytics exports

Current product contract:

• REST returns one candle per request
• JSON-RPC returns one candle per request
• WebSocket streams live candle state after subscribe with wss://api-hyperliquid-index.n.dwellir.com/YOUR_API_KEY; wss://api-hyperliquid-index.n.dwellir.com/YOUR_API_KEY/ws is also accepted
• allCandles WebSocket streams emit arrays of changed candle snapshots and bill by the number of candle objects delivered in each data frame

Sparse Candles

This means:

• the dataset preserves real trading activity instead of synthesizing flat bars
• downstream consumers that need dense chart series can fill missing intervals on their side
• response counting for historical retrieval should assume one response per existing candle, not one response per theoretical time bucket

Response Volume for One Market Over 30 Days

The current OHLCV contract returns one candle per response for REST and JSON-RPC. That means historical retrieval volume is easy to estimate directly from the interval you need.

For example, a full 30-day pull of 1m BTC candles is about 43,200 responses, while the same window at 1s resolution is about 2.59 million responses.

Example Candle

{
  "s": "BTC",
  "i": "1m",
  "t": 1774911600000,
  "T": 1774911659999,
  "o": "66745",
  "h": "66746",
  "l": "66687",
  "c": "66687",
  "v": "4.78586",
  "q": "319316.665",
  "n": 256,
  "x": true
}

Data Quality

• Accuracy: Derived from canonical raw fill history
• Consistency: One schema across all shipped intervals and transports
• Recency: Historical backfill plus live candle upkeep

Note on Sweep Executions

In fast markets, one large order can trade through several price levels in a single sweep. When candles are derived from raw fill data, those exact match prices are preserved, which can occasionally produce slightly different highs or lows than a venue's own public candlestick feed.

Think of the book like this:

• ask at 69344
• ask at 69343
• then one aggressive order arrives and buys through both levels

In one BTC 5-minute bar we validated, that sweep matched:

• 0.08487 BTC at 69344
• 0.34059 BTC at 69343

If you keep the exact fill prices, the candle for that slice is:

• low = 69343
• volume = 0.42546 BTC

From our benchmarking against the public Hyperliquid endpoint, it appears that their candlestick feed compresses some sweeps differently. We base our candles directly on node fills and preserve the executed prices instead of re-aggregating that move away, so the same sweep can appear as:

• low = 69344
• volume = 0.42546 BTC

This is intentional. Dwellir candles are built from the underlying node fill stream, so highs, lows, and volume reflect the executed prices we observe rather than a more compressed candle representation. That can produce small high or low differences versus another public feed, but it is the more execution-faithful view of what actually traded.

Related Data Products

Access

Use your Dwellir API key directly against the public OHLCV host, or contact the Dwellir team if you need help with access and product fit.

• Dashboard: dashboard.dwellir.com
• Pricing: dwellir.com/pricing
• Support: support@dwellir.com