Skip to main content

Documentation Index

Fetch the complete documentation index at: https://pmxt.dev/docs/llms.txt

Use this file to discover all available pages before exploring further.

The Router is PMXT’s cross-venue intelligence layer. One API key gives you a unified view of every prediction market — Polymarket, Kalshi, Limitless, and 8 more venues — through a single interface. 
import pmxt

router = pmxt.Router(pmxt_api_key="pmxt_live_...")

# Search every venue at once
markets = router.fetch_markets(query="fed rate cut", limit=5)

# Find the same market on other venues
matches = router.fetch_market_matches(market_id=markets[0].market_id)

# Compare prices across venues
prices = router.compare_market_prices(market_id=markets[0].market_id)
The Router only needs a PMXT API key — venue credentials are not required for search, matching, or price analysis.

​
What you can do

Cross-venue search

Search markets and events across every venue in a single query. Results come from a shared catalog — ~10ms, not 11 sequential API calls.

Market matching

Given a market on any venue, find the same or related market on every other venue. Each match has a relation type, confidence score, and reasoning.

Price comparison

Side-by-side bid/ask for the same market on every venue. See where it’s cheapest to buy and where it pays the most to sell.

Related markets

Find markets whose resolution conditions are subsets or supersets of a target market — narrower or broader markets on other venues.

Matched markets

Compare prices for the same market across venues. See price differences at a glance.

Compose with venues

Use the Router for data, venue exchanges for trading. Same SDK, same schema — marketId works everywhere.

​
How it works

The Router is backed by a continuously-updated Postgres catalog that ingests markets, events, and outcomes from every venue PMXT supports. When you search or match, you’re querying this catalog — not fanning out to 11 different APIs. Matching uses embedding-based similarity and LLM-verified relation classification to find semantically equivalent markets across venues, even when venues use different titles, different bracket ranges, or different framing.
Venue A (Polymarket)Venue B (Kalshi)Relation
”Will BTC exceed $100k by Dec?""Bitcoin above $100,000 on Dec 31”identity
”Democrats win presidency""Democratic nominee wins popular vote”subset
”Fed cuts rates in 2025""Fed cuts rates in June”superset

​
Get started

  1. Get an API key at pmxt.dev/dashboard. It starts with pmxt_live_ and works immediately.
  2. Install the SDK: 
pip install pmxt
  1. Create a Router and start querying:
import pmxt

router = pmxt.Router(pmxt_api_key="pmxt_live_...")
markets = router.fetch_markets(query="election")
That’s it. Everything the Router returns uses the same unified schema as every other PMXT exchange, so you can pass a marketId from the Router straight into a venue exchange when you’re ready to trade.
Order routing is coming. Today the Router is read-only. Smart order routing (best-execution across venues) is on the roadmap.