TOTAL VOLUME:
$164.7b
24H VOL:
$148,169,742
24H TRANSACTIONS:
3,258,908,236
OPEN INTEREST:
$1,658,654,544
346,710
Markets across
38,022
events
MATCHED EVENTS:
3,752
PLATFORM COVERAGE:
5
Polymarket:
36%
VS.
Kalshi:
64%
Polymarket wallets are public, so tools can mirror them. Here is how a copy trade works, what it costs, which platforms allow it, and the documented bot scams.
Jared Polites
Nov 13, 2026

TL;DR
Have you ever looked at a Polymarket leaderboard and wondered what it would be like to just do what the top wallet does? That is a Polymarket copy trade, and it is possible because the platform's trades are public.
This guide explains how it works, what it costs, which platforms allow it, and the risks that matter most. It uses public prediction market data and Polymarket's documentation, and it is not advice to copy anyone.
It is a position you open because another wallet opened it. A tool watches the public history of a wallet you choose. When that wallet buys or sells a contract, the tool places a similar order from your account.
Nothing about this is built into Polymarket itself. The copying is done by outside software that reads public data and then trades through Polymarket's API.
You can also copy a trade by hand. You look at a wallet's recent activity and decide whether to take the same position yourself. Bots just do it faster and without a human in the loop.
Three steps happen every time. A tool reads a wallet's activity, decides whether to match it, and signs an order from your account.
A lot. Polymarket's Data API returns positions, activity and statistics for any wallet address, with no login. The docs describe it as a way to "understand an account's current exposure and trace how its activity has changed over time."
The main reads are open and closed positions, activity, portfolio value, profit and loss history and volume. A separate leaderboard route ranks traders by profit and loss or volume over a day, week, month or all time.
This wallet-level data has limits. Deposits and withdrawals are hidden by default. Profit and loss points are cumulative, so you can't add them up as period returns. A null profit and loss value means unavailable, not zero. Our Polymarket API guide covers the endpoints in detail.
By signing orders on your behalf. Trading on Polymarket needs a signer, and every order is signed before it reaches the order book. A copy tool needs some form of that access.
The safest route Polymarket documents is a session key. It is a separate signing key that a Deposit Wallet owner authorizes to trade. The docs say a session key "cannot withdraw funds from the Deposit Wallet." It is a beta feature, authorizations last 180 days, and the owner can revoke one early.
The riskier route is handing over the wallet's private key. We cover why that matters below.
We didn't find one. Polymarket's documentation describes an API for programmatic trading. It does not describe a feature for following other traders, and we found no Polymarket guidance on third-party copy tools. What its terms say about bots is not publicly documented in the pages we could read.
Every copy tool we saw is run by an outside company or developer. We don't name them here. Most pages promoting them are written by the tools' own makers, and we don't rank or endorse any of them.
A Polymarket bot is any program that reads market data or places orders through the API. Copying a wallet is one use. Others include quoting both sides of a market, watching for price gaps and rebalancing positions.
Polymarket publishes a trading API, official software kits and published rate limits. The platform's central limit order book accepts orders signed by a wallet or a session key. For the mechanics, see our guide to the Polymarket API.
Most bot advice online is written by people selling something. Run the numbers on fees and slippage before you assume any automated approach pays.
Not in the same way, because there is no wallet to follow. Kalshi settles trades on its own ledger, and its public trades feed lists price, size and side with no account identifier. Our earlier research found the same for Polymarket US, which requires full identity verification and shows no public wallet history.
Bots still exist on both. Kalshi publishes a trading API and a demo environment with mock funds, which is where developers test bots before trading real money. Our Kalshi API guide covers signing, limits and the demo.
Polymarket's U.S. and international products also differ in who can use them. See our guide to Polymarket in the U.S. for that split.
Only on-chain platforms publish wallets. This table comes from our earlier research on position transparency and wasn't re-verified for this page.
| Platform | Wallet history public? | Settlement |
|---|---|---|
| Polymarket (international) | Yes | On-chain |
| Polymarket US | No | KYC accounts, regulated brokers |
| Kalshi | No | Internal ledger |
| Limitless | Yes | On-chain (Base) |
| Predict.Fun | Yes | On-chain (BNB Chain) |
| Opinion | Yes | On-chain (BNB Chain) |
Limitless has the same wallet-level visibility as Polymarket but less liquidity, so patterns take longer to build.
Predict.Fun and Opinion are both on BNB Chain. A public wallet tells you about a trader's history. It tells you nothing about whether the platform is regulated. Our guide to prediction market whales covers that difference.
More than the original trade, in three ways. You pay a fee, you cross the spread, and the price may move before your order fills.
Polymarket charges only takers. The formula is fee = shares x fee rate x price x (1 minus price). Politics markets use a rate of 0.04, crypto 0.07 and sports 0.05, and geopolitics markets are fee-free. Makers pay nothing.
A copy bot that takes the best available price is a taker. On a hypothetical 100-share copy at 60 cents in a politics market, the fee is 100 x 0.04 x 0.60 x 0.40, or $0.96. See our guide to prediction market fees for each platform.
Slippage works against you too. If the wallet bought at 60 cents and the price is 63 cents when your order fills, a contract that pays $1 now returns 37 cents, not 40. That is a hypothetical, and it is the kind of gap that follows any trade you see only after it happens.
Copying carries the original trader's losses as well as the gains, and it adds risks of its own.
We found no rigorous published research on whether copying prediction market wallets is profitable. Treat any claim of returns as unproven.
Yes, in documented cases. Two campaigns in 2026 targeted people who wanted to run a Polymarket bot.
On May 20, 2026, nine npm packages posing as Polymarket trading tools were published within about two minutes, according to the OSV advisory. One, polymarket-bot, sends Ethereum private keys to an attacker's server. It reads any PRIVATE_KEY value from the project's .env file. It wraps real Polymarket software so the commands appear to work.
On June 30, 2026, the security firm SafeDep reported a fake arbitrage bot repository with 36 stars and 53 forks. The bot's setup asked for a private key in a .env file. A dependency called clob-client-math ran on install and, per SafeDep, targeted wallet vaults, browser passwords, SSH keys and cloud credentials. SafeDep tied 30 packages to ten npm accounts.
Stars and forks can make a repository look established. The people affected did not search for an obscure package. They cloned something that looked like a working bot.
Don't give any tool your main wallet's private key. Polymarket's docs show examples that load a signer key from environment variables, which is the same place the malware looked.
These steps reduce the exposure:
Revoking a session key stops further trading and cancels open orders, but the docs say full revocation can take several minutes.
Use the data as research, not as an instruction. A wallet's history shows what it did. Cross-platform prices show what markets collectively think now.
PredictionHero tracks all five platforms. Our guide to where to find prediction market data explains each source and its limits.
For a price-gap view, see our piece on prediction market arbitrage.
It means a tool watches a public Polymarket wallet and places matching orders from your account. It works because the international platform publishes every wallet's positions and activity. The copying is done by third-party software, not by Polymarket, and you pay the fees and price moves on your own orders.
We found none. Polymarket's documentation describes an API for programmatic trading and says nothing about following other traders. Every copy tool we saw is run by an outside developer. Polymarket's guidance on third-party tools is not publicly documented in the pages we reviewed, so check its current terms.
Unknown. We found no rigorous published research on returns from copying prediction market wallets. Copies pay taker fees, cross the spread and arrive after the original trade. Leaderboards also show survivors, which makes a ranking weak evidence of skill. Treat any promised return as unproven.
Not by wallet. Kalshi settles on its own ledger, and its public trades list price, size and side with no account identifier. Developers do build bots on its trading API, and Kalshi offers a demo environment with mock funds for testing. Those bots follow rules you write, not another trader.
Some are not. In 2026, npm packages and a fake arbitrage bot repository were reported to steal wallet keys. Use a session key that cannot withdraw, fund the wallet lightly and read the code. Our guide draws on public prediction market data and security advisories.
PredictionHero aggregates publicly available prediction market data for informational purposes only. This is not financial advice. Prediction markets may not be available in all jurisdictions.
Related

A developer's map of the Polymarket API: Gamma, CLOB and Data endpoints, authentication, WebSocket feeds, fees and rate limits, checked October 2026.
Jared Polites · Oct 25

Position size alone proves nothing. Here's how to separate real information from ordinary risk-taking and wash trading, and why some platforms let you check a wallet's history while others don't.
Jared Polites · Sep 4

Kalshi, Polymarket, Predict.Fun, and Opinion now all charge takers a fee that peaks on a 50-cent contract and fades toward the extremes.
Jared Polites · Aug 18