PredictionToolkit logoPredictionToolkit
Language: English
Search tools
← Historical data

Dataset

PolyOrderbooks: compare depth-based fills with your crypto rules

PolyOrderbooks offers Polymarket crypto and sports history through REST, Python and MCP, alongside a web-based Backtest AI product. Data queries and crypto strategy replay have separate plan choices and coverage limits.

3.5/ 5Editorial ratingEditorial assessment
Open PolyOrderbookspolyorderbooks.com
Available on
Web · API · MCP

Screenshots and interface

Product preview1 / 2

Historical-book product and provider example.

What is PolyOrderbooks?

Books, prices and liquidity metrics share a timeline. Both outcome books expose resting price and size levels, so researchers can compare a midpoint assumption with fills produced by walking depth. The finest documented query interval is 250 ms on every plan, including Starter.

Backtest AI turns a supported crypto description into entry and exit rules, then reports slippage, partial fills, exit reasons, P&L and drawdown. Review the parsed rules before reading results. Sports books belong to the data service; sports strategies are outside the documented AI replay, and sports history stays capped at 30 days.

Who it is for

  • Researchers comparing midpoint and depth-based fills.
  • Developers using crypto and sports history.
  • Users testing crypto Up/Down entries and exits.

Features and coverage

  • 250 ms L2 books

    Retrieve price and size levels for both outcomes.

  • Aligned history

    Query prices and market metrics over the same interval.

  • Resolved market discovery

    Find archived markets and settlement outcomes.

  • Natural-language backtests

    Convert supported crypto strategies into executable entry and exit rules.

  • Execution reports

    Inspect slippage, partial fills, exits, P&L and drawdown.

  • Exportable runs

    Export trade rows as CSV and reports as JSON.

  • REST, Python and MCP

    Use REST, the Python SDK or the official MCP server.

Workflows with PolyOrderbooks

Test a crypto rule

Set asset, duration, entry price and exits, review the parsed rules and run a bounded replay before changing parameters.

Measure fill sensitivity

Walk historical book levels at different sizes and compare with the midpoint series.

Build an assistant research client

Connect the MCP server with a data key and request one narrow market/time slice.

Pricing and total cost

Every plan includes the documented 250 ms resolution. Paid data tiers change history and request capacity; Backtest AI is separate. Sports retains only the latest 30 days.

Starter

Free

  • 3 days of history; 250 ms resolution.
  • 60 requests/minute; 1,000/day.
  • 1 AI backtest and 3 strategy backtests.

Data

US$19 / US$29 / US$39 / US$49 per month

  • 30 / 60 / 90 / 120-day history windows.
  • Base speed: 300 requests/minute; 50,000/day.
  • Sports: latest 30 days; longer windows extend crypto only.

Backtest AI

US$29/month standalone or +US$19/month

  • 500 AI backtests/month.
  • Unlimited strategy backtests.
  • Crypto Up/Down replay; sports books remain in data plans.

Fees and usage costs

  • Fast speed adds US$19/month; Heavy adds US$39/month.

Getting started

Before you start: Sports history is 30 days; longer windows apply to crypto.

  1. Create an account, verify email and generate an API key.
  2. Choose a market and bounded time range; retrieve books and follow cursors.
  3. For Backtest AI, review parsed entries and exits and export the run for comparison.

Sports history stays limited to 30 days even on longer crypto plans.

Read PolyOrderbooks documentation

Useful links

What to check

  • Sports history is 30 days; longer windows apply to crypto.
  • Backtest AI supports crypto Up/Down, without sports strategy replay.
  • 250 ms snapshots do not establish every event’s order within the interval.
  • 250 ms is a documented capture/query interval, not independently measured end-to-end latency.
  • Speed upgrades and longer history are separate choices.

A closer look

Ratings & reviews

Editorial rating
3.5/ 5
0 user reviews
PredictionToolkit editorial

Our take on PolyOrderbooks

Fine-grained books and distinct data/backtest products make research scope clear. Sports retention, crypto-only AI replay and separate upgrades constrain some workflows.

Based on documented features, setup and access terms. How we rate

User reviews 0

The main score switches to user ratings after 5 published reviews.

Be the first to share your experience

A specific example helps the next person choose the right tool.

Related tools

Browse category
DepthFeed v3 API introduction and dataset scope
DepthFeed
Data API4.0Editorial

Retrieve recorded book ladders, sports mappings and aligned price bars for research and replay.

  • Access API · MCP

Another historical-depth API.

Marketlens Python installation and Strategy quickstart
Marketlens Python SDK
SDK4.0Editorial

Compare fast bar experiments with detailed L2 replay, reusing downloaded history between strategy changes.

  • Language Python

Programmatic Python strategy replay.

Pendulum Flow public archive introduction and capture eras
Pendulum Flow Orderbook Archive
Dataset4.0Editorial

Inspect hourly Parquet files and coverage audits before reconstructing historical Polymarket depth.

Raw capture files for custom reconstruction.

Frequently asked questions about PolyOrderbooks

Does Starter include books?

Yes, L2 at 250 ms, with smaller history and request limits.

Does a 120-day plan give 120 sports days?

No. Sports remains limited to 30 days.

Can I inspect the generated rules?

Review the parsed entry and exit rules before interpreting the report.

Can I export results?

Trades export as CSV and reports as JSON.