Open data for prediction market research.

The problem

Most of the data is inaccessible.

Prediction markets are the most efficient public mechanism for pricing uncertainty. Every day, hundreds of millions of dollars move through Polymarket, Kalshi, and Hyperliquid on questions ranging from central bank rate decisions to geopolitical conflict resolution. The data these markets generate — real-time orderbooks, price histories, volume profiles, cross-venue spreads — is rich. Most of it is inaccessible to the people who would do the most interesting work with it.

Academic researchers, graduate students, independent quants, and open-source builders generally cannot afford the infrastructure to collect, normalize, and store tick-level orderbook data across multiple prediction market venues. Even when they can, the lack of structured index-level abstractions (baskets, compositions, NAV series) means they are working with raw ingredients rather than the financial objects that institutional analysis requires.

What the grant provides

23 endpoints. Same data plane as the institutional dashboard.

The Itô Research Grant gives accepted researchers free, authenticated access to the full Itô public API spanning market data, basket analytics, and hosted backtesting infrastructure. This is not a toy sandbox.

Market and price data

Historical L2 orderbook snapshots across Polymarket, Kalshi, and Hyperliquid, sourced from Itô's ClickHouse hot storage, queryable in 24-hour windows up to 10,000 rows per request. Bulk daily close prices for up to 50 markets per request, with up to 365 days of lookback. Individual market detail including categorization, basket assignment, geographic scope, volume, and platform metadata. Market price time series: daily close prices with volume, up to a year of history.

Basket-level analytics

Itô constructs systematic, transparent indices (“baskets”) over prediction market contracts. Grant recipients get full access to the basket analytics surface: all live indices with current NAV and 24-hour change, constituent markets with sides, weights, marks, and historical composition snapshots, time-ordered NAV series from intraday to all-history, 1-day/1-week/1-month returns plus 30-day volatility, and daily volume aggregated from the platform trade ledger.

Hosted backtesting (ATOP)

Access to Itô's ATOP (Algorithmic Trading on Prediction markets) backtesting infrastructure. Browse strategy templates running on a Rust hot-path runtime, create custom strategies with custom parameters and venue configurations, explore replay datasets, validate and plan multi-window train/test experiments, and submit backtest jobs to retrieve result artifacts (PnL, fill counts, roll summaries).

Who should apply

Researchers, students, quants, builders.

Academic researchers studying prediction market microstructure, information aggregation, or event-driven pricing. Graduate students writing theses on market efficiency, cross-venue arbitrage, or portfolio construction over binary contracts. Quantitative analysts building systematic strategies who need historical orderbook and price data to backtest against. Open-source builders developing tools, libraries, or dashboards that advance the prediction market ecosystem.

Research areas of interest

What we want to see.

Cross-venue price discovery (Polymarket vs. Kalshi vs. Hyperliquid)

Basket construction methodologies and index rebalancing strategies

Volatility estimation and risk factor models for prediction markets

Market-neutral and delta-hedged execution strategies

Information aggregation efficiency in thin vs. liquid markets

API access

Rate limits and scopes.

Grant recipients receive an API key scoped to baskets:read, markets:read, backtests:read, and backtests:write. Read requests are limited to 120 per minute; write requests to 10 per minute. Authentication is a Bearer token in the Authorization header. Full API reference documentation is available at itomarkets.com/api-docs.

Future grants

Where this is going.

The research program will grow as Itô grows. Expanded data coverage with additional venues, higher-frequency snapshots, and historical depth beyond the current hot storage window. Streaming data access with real-time SSE price streams for live research and training pipelines. Compute grants with dedicated backtesting budgets for longer replay windows. Publication support including co-authored papers, featured research on Itô Research (Substack), and presentation slots at Itô-hosted events.

As Itô expands into oil, gold, supply-chain, and real-world exposure baskets, grant recipients will be among the first to access data from these new asset classes.

Apply

How to get access.

Send an email to alejandro@itomarkets.com with your name, affiliation, and role. Include a brief research proposal (1–2 paragraphs) describing what you plan to study or build, which data surfaces you need (market data, basket analytics, backtesting, or all three), your expected timeline, and any prior work (papers, repos, projects).

Applications are reviewed on a rolling basis. Accepted researchers will receive an API key within one week of approval.

Attribution

Data sources.

Market data served through the Itô API is sourced from the following venues and libraries. Grant recipients should include appropriate attribution when publishing research.

“pmxt” (SDK) by pmxt.dev, sourced from github.com/pmxt-dev/pmxt, used under MIT License.

“pmxt” (data) by pmxt.dev, sourced from github.com/pmxt-dev/pmxt, used under CC BY 4.0.

“Hyperliquid Python SDK” by Hyperliquid, sourced from github.com/hyperliquid-dex/hyperliquid-python-sdk, used under MIT License.

“Hyperliquid SDK” by QuickNode, sourced from github.com/quiknode-labs/hyperliquid-sdk, used under MIT License.

“Polymarket CLOB API” by Polymarket, sourced from polymarket.com. Subject to Polymarket's Terms of Service.

“Kalshi API” by Kalshi, sourced from kalshi.com. Subject to Kalshi's Terms of Service and data usage policies.

“Hyperliquid API” (data) by Hyperliquid, sourced from hyperliquid.xyz. Subject to Hyperliquid's Terms of Service.

Grant recipients are responsible for complying with each source's license and terms of service when using or redistributing data obtained through the Itô API.