Exchange data hub
Download Lighter historical perpetual order books, tick trades, open interest, and funding data in normalized hourly Parquet files.
Research Lighter perpetual markets without first building a historical WebSocket archive. Files retain the venue context needed to reason about market IDs, nonces, funding, and special trade events.
Coverage
Each row is a separate landing page with its own exchange IDs, schema, sample path, caveats, and documented history start. A listed market applies only to that dataset row.
| Dataset | Stored markets | Archive ID | History start | Example object path |
|---|---|---|---|---|
| Order Book | Perpetuals | lighter | 2025-12-30 | lighter/2025-12-31/20/BTC_orderbook.parquetHourly Parquet/Zstd |
| Tick Trades | Perpetuals | lighter | 2025-12-30 | lighter/2025-12-31/20/BTC_trades.parquetHourly Parquet/Zstd |
| Ticker Statistics | Perpetuals | lighter | 2025-12-30 | lighter/2025-12-31/20/BTC_ticker.parquetHourly Parquet/Zstd |
| Open Interest | Perpetuals | lighter | 2025-12-30 | lighter/2025-12-31/20/BTC_open_interest.parquetHourly Parquet/Zstd |
| Funding & Mark Price | Perpetuals | lighter | 2025-12-30 | lighter/2025-12-31/20/BTC_mark_price.parquetHourly Parquet/Zstd |
Before analysis
Normalization makes columns consistent; it does not erase contract design, symbol mapping, or limits in the exchange's public feed.
01
Lighter's API addresses books by numeric market index; the archive exposes display symbols but researchers should retain the market mapping used for the requested period.
02
Order book continuity is checked with venue nonces. API-server offsets can change after reconnects and are not a substitute for matching-engine nonce continuity.
03
The venue can publish deleverage trade events with special price semantics. Do not silently treat every trade row as an ordinary buyer-seller match.
Next steps
Use the dataset pages for exact semantics, then download a small interval and validate it against your research assumptions.
