Prediction Market Scanner — Polymarket vs Kalshi 📊

Search any topic and get the live binary (yes/no) markets from both Polymarket and Kalshi in one dataset — side by side, with the price gap between the two platforms.

Why compare the two platforms?

Polymarket and Kalshi are the two biggest prediction markets, and they price the same real-world questions independently — "Will there be a government shutdown?", "Will X win the election?", "Who wins the Nobel Peace Prize?". When they disagree, that gap is a signal: an information edge, a liquidity difference, or a mispricing worth a closer look.

Doing this by hand means opening both sites, hunting for the matching market on each, and eyeballing the odds. This actor does it in one query: you give it a topic, it pulls the live markets from both platforms, pairs the ones asking the same yes/no question, and shows you exactly how far apart the two platforms are.

What you get

Two kinds of rows in one dataset:

versus rows — the same yes/no question found on both platforms:

Field	Description
polymarket_title / kalshi_title	The matched market on each platform
polymarket_yes / kalshi_yes	Live YES price on each platform (0–1)
yes_gap_pts	How far apart the two YES prices are, in percentage points
polymarket_no / kalshi_no	Live NO price on each platform
similarity	How confidently the two titles were matched (0–1)
polymarket_volume / kalshi_volume	Traded volume on each side
polymarket_url / kalshi_url	Direct links to both markets

single rows — markets found on only one platform (full topic coverage, so nothing is hidden). Set include_unpaired: false if you only want the cross-platform comparisons.

Rows are sorted by the biggest YES-price gap first — the most interesting divergences on top.

Zero markets on both platforms = never billed. You only pay for runs that actually return data.

How the matching works — read this

Markets are paired by title similarity, not by fuzzy price guessing. Two markets are only compared if they genuinely ask the same yes/no question (a strict default similarity of 0.62, tunable). Multi-outcome markets (e.g. Kalshi's "Who will win the Nobel Peace Prize?") are split by candidate so "Zelenskyy" is compared to "Zelenskyy", never to the wrong person. Resolved and illiquid markets (prices stuck at 0 or 1) are dropped so you never see a fake gap.

This is honest, mechanical matching. The similarity score is shown on every pair so you stay in control — raise min_similarity for stricter pairing, lower it for wider coverage.

Sources

Public, login-free read APIs (no keys required):

• Polymarket — Gamma public search + live outcome prices.
• Kalshi — public markets API, matched by topic across its market series.

Use cases

• Cross-platform edge hunting — spot where the two markets disagree on the same event.
• Odds monitoring — track a topic (election, macro event, award) across both platforms daily.
• Research & journalism — cite what prediction markets currently imply, from both venues at once.
• Model inputs — feed consolidated yes/no probabilities into your own analysis.
• Liquidity comparison — see which platform has more volume on a given question.

Input

Field	Type	Default	Description
query	string	"government shutdown"	Topic to scan (e.g. "presidential election", "recession", "nobel peace prize")
platforms	array	["polymarket", "kalshi"]	Platforms to compare (both required for versus rows)
min_similarity	number	0.62	How similar two titles must be to be paired (0–1)
min_gap_pts	number	0	Only report pairs whose YES prices differ by at least this many points
include_unpaired	boolean	true	Also return single-platform markets (full topic data)
max_results	integer	50	Cap on rows returned
max_markets_per_platform	integer	100	Markets pulled from each platform before pairing
request_timeout_secs	integer	30	HTTP timeout in seconds

Example input

{
  "query": "nobel peace prize",
  "min_similarity": 0.62,
  "min_gap_pts": 2,
  "include_unpaired": false
}

Example output (versus row)

{
  "match_type": "versus",
  "query": "nobel peace prize",
  "similarity": 0.701,
  "yes_gap_pts": 2.6,
  "polymarket_title": "Will Pope Leo XIV win the Nobel Peace Prize in 2026?",
  "kalshi_title": "Who will win the Nobel Peace Prize? — Pope Leo XIV",
  "polymarket_yes": 0.044,
  "kalshi_yes": 0.07,
  "polymarket_no": 0.956,
  "kalshi_no": 0.93,
  "polymarket_volume": 18234.0,
  "kalshi_volume": 5120.0,
  "polymarket_url": "https://polymarket.com/event/...",
  "kalshi_url": "https://kalshi.com/markets/..."
}

Tips

• Start broad (min_gap_pts: 0) to see everything, then raise it to surface only real divergences.
• If you get too many loose pairs, raise min_similarity toward 0.7+.
• Kalshi has no text search, so coverage depends on the topic matching a Kalshi market series — political, economic, crypto and world-event topics have the best overlap with Polymarket.

Disclaimer

This actor returns live public market data for informational purposes only. It is not financial, investment or trading advice, and a reported price gap is not a guaranteed profit — prices move, liquidity varies, and matched markets may differ in resolution details or timing. Always read both markets' rules before acting. You are responsible for your own decisions and for complying with the terms of each platform and your local laws.