Derivatives data,
unfragmented.
Prediction markets and crypto options in one normalized schema. Tick-level trades, L2 orderbooks, OHLCV, and Black-76 greeks across Polymarket, Kalshi, Deribit, Lyra, OKX, Aevo, and Delta — pay only for what you download.
Built for the way quants actually work.
Every tick from every venue — cleaned, normalized, and stored, so you don't have to. What used to take a quarter of data engineering now takes a single API call.
Prediction markets
Polymarket + Kalshi — the two biggest event-market venues in one schema. Trades, L2 orderbooks, OHLCV candles, market stats, and per-wallet daily analytics.
Crypto options + futures
Deribit, Lyra (Derive), OKX, Aevo, and Delta. Full options-chain trades plus L2 snapshots, with Black-76 greeks computed at trade time.
Bulk Parquet downloads
Pre-aggregated daily Parquet served via presigned R2 URLs. pandas, polars, and DuckDB read them directly — no proprietary format, no lock-in.
Backtest-ready
Point-in-time accurate. Every row timestamped at venue ingest — no look-ahead bias, no rolled-up artifacts, no silent revisions.
REST API + catalogue
Catalogue series across every venue with one endpoint. Estimate the cost before you pay, then pull the data with a single X-API-Key header.
Pay-as-you-go
Trades $3 · orderbook $5 · snapshots $8 · greeks $12 · OHLCV $20, per compressed MB. $50 minimum order, no subscriptions, no seats.
From catalogue to your pipeline.
No scraping chains or exchange APIs. No schema reconciliation. Catalogue a series, estimate the cost, download the Parquet — with the tools your team already uses.
# 1 · find a series in the catalogue curl "…/v1/series?venue=deribit&type=trades" # 2 · estimate the cost (free) curl "…/v1/series/deribit:btc-options/estimate" # 3 · download the Parquet (presigned R2) curl -OJ "…/v1/series/deribit:btc-options/download" # → read it anywhere import polars as pl df = pl.read_parquet("deribit_btc_options_2026-05-24.parquet")
Every venue, one schema.
*“Since” is the earliest date for which we hold data at a venue. It is not a guarantee of continuous coverage: not all data types are available from that date, availability varies by data type and instrument, and history is continuously backfilled. Always confirm the exact available range for a given series in the catalogue before purchasing — quoted figures are provided as-is, without warranty.
Pay per MB, not per seat.
Five data types, every venue. Billed by compressed size — $50 minimum order, no subscription, no seats.
Browse the catalogue free · no card to explore
Questions, answered straight.
Still curious? Email support@supagamma.com and a real engineer will reply.
What is SupaGamma?
SupaGamma is a historical data platform for Polymarket prediction markets. We continuously archive orderbook snapshots and on-chain trade data so you can backtest strategies, build models, and conduct research.
How do I access the data?
Sign up, generate an API key, and start downloading via our REST API. We support CSV and JSON export formats. Use any HTTP client — Python, JavaScript, curl, or any language that can make HTTP requests.
What data do you provide?
We provide two core datasets: orderbook snapshots (full depth-of-book captured every 60 seconds across 1,000+ markets) and on-chain trade history (every fill indexed directly from the Polygon blockchain).
How much does it cost?
Per-data-type pricing: Trades $3/MB, OHLCV candles $20/MB, Orderbook snapshots $5/MB. $50 minimum order, no subscriptions. You only pay for what you download.
How far back does the data go?
Our historical archive covers Polymarket data from early 2024 onwards, with continuous backfill expanding coverage. New orderbook snapshots and trades are captured in real-time and added to the archive daily.
Stop scraping. Start querying.
Browse the full catalogue for free. Pay only for the data you download — from $3/MB, $50 minimum, no subscription.