Kalshi fees explained, with a fee calculator

Kalshi charges takers 1.75 cents per contract at a price of 50 cents. At 5 or 95 cents the formula gives about 0.33 cents per contract, before Kalshi rounds the fee up. Makers pay nothing on most markets, and there is no fee at settlement, when a market closes and pays out.

The taker fee is 0.07 times the number of contracts times P times (1 minus P), rounded up. P is the price in dollars. A taker is an order that trades at once with an order that is already waiting. A maker order waits in the order book until someone trades with it.

Deposits and withdrawals by ACH (Automated Clearing House, the standard US bank transfer) are free (fee schedule effective 07/07/2026, checked 10/10/2026).

In June I pulled Kalshi’s series data to compare exchanges for a few thin-edge ideas, where fees can wipe out a small advantage. The fee type said quadratic, a curve that rises and then falls with the price. That changed which idea I ranked first.

How much does Kalshi charge per trade?

Kalshi charges a trading fee only when your order matches one that is already waiting, which makes you the taker. On a 100-contract order at 50 cents, the taker fee is $1.75. At 5 cents the same 100 contracts cost $0.34 in fees, and at 95 cents it is also $0.34.

Those numbers come straight from the table in Kalshi’s fee schedule, which I read on 10/10/2026. Kalshi lists no membership fee and no settlement fee, and canceling a waiting maker order is free. So the only fee I model in my own work is the one charged on the trade itself.

How are Kalshi fees calculated?

Kalshi publishes two formulas, one for takers and one for makers. A series is a group of related markets, and Kalshi sets a multiplier, M, for each one.

taker fee = round up( M x 0.07   x C x P x (1 - P) )
maker fee = round up( M x 0.0175 x C x P x (1 - P) )

P = contract price in dollars
C = number of contracts
M = series multiplier (taker default 1, maker default 0)

Because the maker default is 0, most maker orders pay nothing. Kalshi keeps a list of series with their own multipliers, and a handful of them charge nothing on either side. A long list of sports, economics and award markets charge makers at a multiplier of 1. Combo markets bundle several outcomes into one contract, and there the maker multiplier is 2, so makers pay half the taker rate.

For one contract at 50 cents, the formula says 1.75 cents. The table in Kalshi’s fee schedule charges 2 cents. My calculator follows the table, so it shows the higher number.

A short note on rounding. Kalshi’s API (application programming interface) docs say Kalshi tracks rounding across each whole order, and some accounts keep their balance to $0.0001. Those accounts get a rebate on overpaid rounding, so a real order can cost a little less than the table says.

The break-even win rate is how often you must be right to lose no money after fees. It assumes you buy at that price and hold to settlement.

Price	Contracts	Fee	Fee as share of cost (after rounding)	Break-even win rate
5 cents	100	$0.34	6.80%	5.34%
25 cents	100	$1.32	5.28%	26.32%
50 cents	100	$1.75	3.50%	51.75%
50 cents	1	$0.02	4.00%	52.00%
90 cents	10	$0.07	0.78%	90.70%
95 cents	100	$0.34	0.36%	95.34%

Why does the Kalshi fee peak at 50 cents?

The fee peaks at 50 cents because P times (1 minus P) is largest when P is one half. The curve is a hill, and both sides match.

The fee tracks how uncertain the contract is, so a coin flip carries the biggest fee. But the hill hides two different costs.

Divide the fee by what you paid and you get 0.07 times (1 minus P). Before rounding, that is 6.65% of your stake at 5 cents. At 95 cents it is 0.35%. Kalshi’s table rounds the fee on 100 contracts at 5 cents up to $0.34, which makes it 6.80%. So longshot buyers, who buy cheap contracts on unlikely events, pay the most as a share of their money.

Now divide the fee by your maximum profit, which is (1 minus P) per contract, and you get 0.07 times P. That is 6.3% of your upside at 90 cents. At 5 cents it is 0.35%. So favorite buyers, who pay up for likely events, give up the most as a share of what they can win.

Neither end is cheap, so no price range on Kalshi is a fee-free zone for a small trader like me.

What win rate do you need to break even after Kalshi fees?

Your break-even win rate is the price plus the fee per contract, as a share of the $1 payout. At 50 cents with 100 contracts that is 51.75%.

This is where fees stopped being a footnote for me. If you think a 50 cent contract is really a 51% event, you still lose money as a taker after fees. Your edge, the gap between your odds and the price, has to clear 1.75 percentage points first.

Longshots are worse on small orders. On one contract at 5 cents, the formula fee is 0.33 cents, which would put your break-even win rate at 5.33%. Kalshi’s table rounds that fee up to 1 cent. So your cost is 6 cents, and you need the event to happen 6% of the time.

The price said 5%. The fee and its rounding together moved your break-even win rate from 5% to 6%, and rounding did about two thirds of that move. On 100 contracts the rounding almost disappears, and the break-even is 5.34%.

Selling early costs you twice, because an exit before settlement is a new trade with its own fee. A round trip at 50 cents is 3.5 cents per contract as a taker. That is before the spread, the gap between the best offer to buy and the best offer to sell.

If you already have your own odds for an event, size the trade with the fee in it. At 50 cents you risk 51.75 cents to win 48.25 cents, so your payout odds are about 0.93 to 1. Enter that with your own win probability in the free Kelly criterion calculator, and it shows how small the position should be.

Does Kalshi charge maker fees?

Kalshi charges maker fees only on series it lists as non-standard. On those, the maker fee is 0.0175 times C times P times (1 minus P). That is a quarter of the taker rate. On every other series the maker multiplier is 0, so maker orders pay nothing.

You can check a series yourself. Kalshi’s API returns each series with a fee_type field and a fee_multiplier field, from GET /series/{series_ticker} (docs checked 10/10/2026). If fee_type says quadratic_with_maker_fees or quadratic_with_combo_maker_fees, makers pay on that series. Plain quadratic means only takers pay.

I check each series before I assume anything. A maker fee applies only when the order fills, and canceling costs nothing.

Does Kalshi charge fees to deposit or withdraw?

Kalshi charges nothing to deposit or withdraw by ACH bank transfer, per its fee schedule. Its help pages on bank and debit card withdrawals both say no fees (checked 10/10/2026). Card deposits cost up to 2%. Crypto deposits and withdrawals can carry fees from the outside company that moves the money.

Kalshi also keeps the right to charge between 0% and 2% on any way you move money. Wire deposits carry no Kalshi fee, but your bank may still charge one.

Why did Kalshi fees change how I designed my bot?

The fee shape made one old idea worth a second look, and then it set hard limits on how I would test it.

Crowds tend to overpay for cheap longshots, and a fade takes the other side of that crowd. In June my notes recorded Polymarket’s fees as a flat percentage by category, while Kalshi’s API returned quadratic. A flat fee kills a strategy that lives at 2 or 5 cents, but a quadratic fee shrinks toward zero there. So I wrote that the longshot fade reopens on Kalshi.

Then I read a paper by Bürgi, Deng and Whelan that studied over 300,000 Kalshi contract prices. Buyers of contracts under 10 cents lost over 60 percent of their money. Across all the contracts each side bought, average returns after fees were -31.46% for takers. For makers they were -11.99%.

My June notes recorded makers as positive. The paper says makers lost money too, only less, so my notes were wrong on that line.

In the thin longshot order books my fade would trade, the fee was small next to the spread. A taker pays that spread and a maker collects it, so any version I would test had to use maker orders.

In July I pulled series data straight from Kalshi’s public API. All 12 of its crypto series that settle every 15 minutes showed the quadratic fee with a multiplier of 1, and none charged makers.

My notes from that check added a warning I still use. A quadratic fee lowers a cost. It does not create an edge. At 5 cents the fee is still 6.65% of the price before rounding, so my edge math simulates the exact fee on every order.

The plan I ranked first used those 15-minute markets. A 10 cent Yes is the same bet as a 90 cent No. The plan used that extreme, a little above the 2 or 5 cents in my June notes. The plan placed a maker order one tick (the smallest price step) inside it, against the overpriced longshot.

It would run only in the final 180 seconds before a market closed. I excluded Bitcoin and Ether because I judged their order books too efficient, which left smaller coins such as Solana and XRP. These are my design settings and nothing more.

The design named one question that could kill it. When my order filled near the close, was the crowd wrong more often than not, or did it know something I did not? The design also requires a stress test that reruns everything as if makers paid the taker fee. If an idea only survives at zero fees, I do not trust it.

Kalshi’s demo exchange, a practice exchange with fake money, did not help. In July it showed 628 of my strategy’s markets open, but only 148 had quotes on both sides. A 48-second dry run of my bot produced zero quotes. So it could test my order code, but it could prove nothing about edge.

As of my July notes, none of this was a tested edge, and no money had gone into it. The next step was a paper test on recorded order book data, to answer that one question about the crowd.

I make no claim here that any of this makes money. What I can say is that reading the fee formula closely changed what I built and what I refused to build.

How do Kalshi fees compare to Polymarket?

For the same trade, 100 contracts at 50 cents, the taker fees land within a cent of each other. The bigger gaps are on the maker side and in access.

Exchange (100 contracts at 50 cents)	Taker fee	Maker side
Kalshi	$1.75	$0 on most series, $0.44 on series with maker fees
International Polymarket, crypto	$1.75	$0, plus a rebate from taker fees
Polymarket US	$1.74	Makers get a $0.31 rebate

International Polymarket is the main exchange at polymarket.com. Its docs describe the same hill-shaped fee as Kalshi’s, with a rate that depends on the category. Crypto is 0.07 and sports is 0.05, politics is 0.04, and geopolitics is 0 (checked 10/10/2026).

Makers never pay there, and they get back a part of the taker fees, 20% on crypto. That corrects my own June notes, so the fee gap is narrower than I thought.

I wrote about my own build there in how I built my Polymarket trading bot. International Polymarket lists the United States as close-only, so US users cannot open new positions there (geoblock page, checked 10/10/2026).

Polymarket US is a separate exchange that the Commodity Futures Trading Commission (CFTC), the US regulator, oversees. Its fee schedule took effect on 10/07/2026. Takers pay at a rate of 0.0695, and makers get a rebate at a rate of 0.0125, rounded to the cent (checked 10/10/2026).

I compare the two exchanges for bot builders in Kalshi vs Polymarket.

If you want to pull fee fields yourself, I am writing a Python tutorial for the Kalshi API next.

This is not financial advice. I am sharing how I read a fee schedule and how it shaped my own design work.

Sources

1. Kalshi, Fee Schedule (last updated and effective 07/07/2026). https://kalshi.com/docs/kalshi-fee-schedule.pdf. Checked 10/10/2026.
2. Kalshi API documentation, fee rounding. https://docs.kalshi.com/getting_started/fee_rounding. Checked 10/10/2026.
3. Kalshi Help Center, bank withdrawals. https://help.kalshi.com/transfer-funds/withdraw-funds/bank-withdrawals. Checked 10/10/2026.
4. Kalshi Help Center, debit card withdrawals. https://help.kalshi.com/transfer-funds/withdraw-funds/debit-card-withdrawals. Checked 10/10/2026.
5. Bürgi, Deng and Whelan, Makers and Takers, The Economics of the Kalshi Prediction Market, UCD Working Paper WP25/19 (July 2025, revised September 2025). https://www.ucd.ie/economics/t4media/WP2025_19.pdf. Checked 10/10/2026.
6. Polymarket documentation, trading fees. https://docs.polymarket.com/trading/fees. Checked 10/10/2026.
7. Polymarket documentation, geographic restrictions. https://docs.polymarket.com/developers/CLOB/geoblock. Checked 10/10/2026.
8. Kalshi API documentation, Get Series (fee_type and fee_multiplier fields). https://docs.kalshi.com/api-reference/market/get-series. Checked 10/10/2026.
9. Polymarket US documentation, fee schedule (effective 10/07/2026). https://docs.polymarket.us/fees.md. Checked 10/10/2026.

Join the trading list if you want the Kalshi API tutorial when it lands.

Keep reading

• How I built my Polymarket trading bot
• Running a production trading bot with Claude Code
• Kelly criterion calculator
• How I build with Claude Code