Kalshi Eye

Kalshi Eye

Historical Kalshi order books, ready for backtesting.

Explore historical bid/ask prices and full-depth order book data across Kalshi markets. Research market behavior, analyze liquidity, and download the data without building your own collection infrastructure.

Explore Market DataView API DocumentationCreate Free Account
KXBTC15M-26OCT091745-45BTC price up in next 15 mins?
recorded sample
Historical YES best bid and ask
Hover or focus the chart for bid/ask

Coverage Oct 9, 2026, 21:31 – 21:44 UTC · includes L2 tape · path step 30s

Sample top-of-book rows from the recorded path
Time (UTC)BidAsk
21:31:300.4100.420
21:34:000.3800.390
21:36:300.3900.400
21:38:300.3000.310
21:40:000.2400.250
21:42:000.2500.260

What we record

Kalshi Eye stores historical market tape for research — not a live trading venue. Indexed catalog size is not the same as markets with recorded order-book history. Gaps can exist where collection was interrupted or a market was never on the recorder.

Sample market
KXBTC15M-26OCT091745-45
Sample coverage
Oct 9, 2026, 21:31 – 21:44 UTC
Path points shown
25 downsampled YES bid/ask observations
L2 on this sample
Yes — 376,754 prints in the tape window
  • Politics
  • Sports
  • Weather
  • Economics
  • Crypto
  • Tech
  • Entertainment
  • World

Categories reflect the exchange catalog we index. Historical order-book availability varies by market and recording window.

View Sample Data

Built for serious prediction market research.

Three ways teams use recorded Kalshi tape — without running their own collector.

  • Backtest trading strategiesEvaluate hypotheses against recorded bid/ask movement, spreads, and available depth. Historical books describe what the market showed — not a guaranteed fill simulator.
  • Analyze market microstructureInspect top-of-book history and, where recorded, full-depth L2 changes — liquidity, spreads, and how the book evolved through the window.
  • Build with historical dataPull files as CSV, Parquet, or JSON, or call the research API from your own stack — pandas, DuckDB, or a custom backtester.

Data layers

Each layer is sized for a job. Snapshots are sampled top-of-book; L2 is the recorded book stream.

  • Top-of-book historyDownsampled best bid and best ask over a window (step_ms). Built for charts and research — not event-by-event exchange messages.
  • Full-depth order bookHistorical L2 anchors and updates when the market was on the recorder. Use this for depth and microstructure work; Full book plan required for L2 exports.
  • Market metadata and outcomesTitles, strikes, open/close times, status, and settlement results alongside the tape when available in the catalog.

Your data. Your workflow.

Browse the archive in the app, download files, or call documented endpoints.

  1. BrowseSearch series and markets, open previews, inspect recorded windows.
  2. DownloadExport snapshots or L2 as CSV, Parquet, or JSON.
  3. APIAuthenticate with a bearer key and pull preview, snapshots, or exports.
Read API DocsExport formats
Production API
curl "https://kalshieye.com/api/v1/markets/KXBTC15M-26OCT091745-45/preview" \
  -H "Authorization: Bearer YOUR_API_KEY"

curl "https://kalshieye.com/api/v1/search?q=bitcoin&limit=20" \
  -H "Authorization: Bearer YOUR_API_KEY"

Start on Explorer

Free Explorer accounts can browse the catalog and market previews, with one free event export. Paid Research and Full book plans unlock broader history exports and L2 — manage plans from your account after you sign in.

  • Explorer$0 — browse + previews + 1 event export
  • ResearchTop-of-book history exports
  • Full bookResearch + L2 / depth exports
Get Started with Kalshi EyeExplore Market Data

Trust and transparency

Kalshi Eye is an independent historical data platform and is not affiliated with Kalshi. Underlying quotes and books originate from Kalshi market data we record.

We record Kalshi market tape into object storage and serve research windows, previews, and exports from that archive.

More detail: data layers · authentication · preview API

Explore Market Data