Free tool, no sign-up
Polymarket Fee Calculator
Polymarket charges the taker rate × P × (1 - P) per contract, and the maker nothing. Because a fixed stake buys more contracts at a lower price, the number you trade against is rate × (1 - P) of your stake, not a flat percentage. On a 4% market that is 2.00% at 50 cents and 3.60% at 10 cents, and the gap decides whether an edge is real.
Work out the fee, and the break-even
A gross edge below 2.00% of your stake loses money on this leg alone. On $100.00 that is $2.00 of fee to clear before any of it is profit. The same rate at $0.90 would need only 0.40%.
Worked example, without JavaScript
The live figures need JavaScript, but everything the calculator computes is also published below as arithmetic and as a full table. A politics market at 4%, a $100 stake and an entry price of $0.50 buys 200 contracts, pays $2.00 in taker fee, costs 1.00 cent per contract, and needs a gross edge above 2.00% of stake to break even. For any other inputs, the break-even table answers directly.
Fee per contract: a parabola
Break-even: a straight line
- 7%, crypto
- 5%, sports, economics, culture, weather, other
- 4%, finance, politics, tech, mentions
Geopolitical markets carry a rate of zero and would sit flat along the bottom of both charts, so they are left off. Makers pay zero in every category, so both charts are taker orders only.
The arithmetic, in full
The published taker fee is charged per contract and depends on that contract's price. Call the contract count C, the price P and the category rate rate:
fee = C × rate × P × (1 - P)
That is the whole formula. No exponent above one, no per-order minimum in the published schedule, and no maker charge: the maker rate is a verified zero, not a discount off the taker rate.
Almost nobody sizes in contracts, though. You decide how much money to risk. Substitute C = stake / P and the price terms collapse:
fee = (stake / P) × rate × P × (1 - P)
= stake × rate × (1 - P)
break-even, as a fraction of stake = rate × (1 - P)
Two different shapes come out of the same formula, and confusing them is where the money goes.
- The fee per contract is a parabola. It peaks at an entry price of 50 cents and falls toward both ends, which is why 10 cents and 90 cents cost the same per contract: 0.36 cents each on a 4% market.
- The break-even as a share of your stake is not a parabola at all. It is a straight line that rises as the contract gets cheaper, because a fixed stake buys nine times as many contracts at 10 cents as at 90 cents. On $100 in a 4% market, the 10 cent trade pays $3.60 and the 90 cent trade pays $0.40. Same fee per contract, nine times the cost as a share of the money at risk.
Every figure on this page follows from those two lines. A number a reader cannot recompute is a number they have to take on trust.
Two trades at the same edge, one of them impossible
Gross edge here means the mispricing you think you have found, as a percentage of stake, before any fee. Two trades at an identical 1.20%:
| Crypto market at $0.8693 | Politics market at $0.2965 | |
|---|---|---|
| Taker rate | 7% | 4% |
| Contracts for $100 | 115.04 | 337.27 |
| Fee per contract | 0.795 cents | 0.834 cents |
| Taker fee | $0.91 | $2.81 |
| Break-even, % of stake | 0.91% | 2.81% |
| Net on a 1.20% gross edge | plus $0.29 | minus $1.61 |
The first trade clears its cost with room to spare. The second cannot win: it needs 2.81% and it has 1.20%, so it is a loss of about $1.61 the moment it fills, before the market moves at all. A single edge threshold cannot tell these two apart. Set the gate at 1.0% and it lets the impossible trade through; set it at 3.0% and it rejects the good one.
Note which way round it falls, because it is the opposite of the intuition. The crypto market is on the highest published rate, 7%, and it is the cheaper trade. The politics market is on the lowest non-zero rate and cannot work. Per contract the two fees are within 5% of each other; as a share of stake they are a factor of three apart. Price dominates rate.
Which is why an entry gate has to be computed per market rather than set
once. The condition is edge% > rate × (1 - P) ×
100, with the rate read from the market being traded.
Break-even by price and category
The full grid, in the page rather than in the tool. Left half is the fee per contract in cents, the parabola; right half is the break-even as a percentage of stake, the straight line. Read across from your entry price to your rate.
| Entry price | 7% cents | 5% cents | 4% cents | 7% break-even | 5% break-even | 4% break-even |
|---|---|---|---|---|---|---|
| $0.05 | 0.33 | 0.24 | 0.19 | 6.65% | 4.75% | 3.80% |
| $0.10 | 0.63 | 0.45 | 0.36 | 6.30% | 4.50% | 3.60% |
| $0.15 | 0.89 | 0.64 | 0.51 | 5.95% | 4.25% | 3.40% |
| $0.20 | 1.12 | 0.80 | 0.64 | 5.60% | 4.00% | 3.20% |
| $0.25 | 1.31 | 0.94 | 0.75 | 5.25% | 3.75% | 3.00% |
| $0.30 | 1.47 | 1.05 | 0.84 | 4.90% | 3.50% | 2.80% |
| $0.35 | 1.59 | 1.14 | 0.91 | 4.55% | 3.25% | 2.60% |
| $0.40 | 1.68 | 1.20 | 0.96 | 4.20% | 3.00% | 2.40% |
| $0.45 | 1.73 | 1.24 | 0.99 | 3.85% | 2.75% | 2.20% |
| $0.50 | 1.75 | 1.25 | 1.00 | 3.50% | 2.50% | 2.00% |
| $0.55 | 1.73 | 1.24 | 0.99 | 3.15% | 2.25% | 1.80% |
| $0.60 | 1.68 | 1.20 | 0.96 | 2.80% | 2.00% | 1.60% |
| $0.65 | 1.59 | 1.14 | 0.91 | 2.45% | 1.75% | 1.40% |
| $0.70 | 1.47 | 1.05 | 0.84 | 2.10% | 1.50% | 1.20% |
| $0.75 | 1.31 | 0.94 | 0.75 | 1.75% | 1.25% | 1.00% |
| $0.80 | 1.12 | 0.80 | 0.64 | 1.40% | 1.00% | 0.80% |
| $0.85 | 0.89 | 0.64 | 0.51 | 1.05% | 0.75% | 0.60% |
| $0.90 | 0.63 | 0.45 | 0.36 | 0.70% | 0.50% | 0.40% |
| $0.95 | 0.33 | 0.24 | 0.19 | 0.35% | 0.25% | 0.20% |
Read the right half top to bottom and the span is the finding: on a 4% market the break-even runs from 3.80% of stake at 5 cents to 0.20% at 95 cents, a factor of nineteen. Across the band where most liquidity sits, 20 cents to 80 cents, it is still a factor of four. A flat threshold over that span is not a conservative simplification. It is wrong at both ends.
The rates, and where they come from
| Category | Taker rate | Maker rate |
|---|---|---|
| Crypto | 7% | 0% |
| Sports | 5% | 0% |
| Economics | 5% | 0% |
| Culture | 5% | 0% |
| Weather | 5% | 0% |
| Other | 5% | 0% |
| Finance | 4% | 0% |
| Politics | 4% | 0% |
| Tech | 4% | 0% |
| Mentions | 4% | 0% |
| Geopolitical | 0% | 0% |
That table is the fallback. The authoritative rate is per market and
published live: Polymarket's Gamma API exposes a feeSchedule
object on each market, and feeSchedule.rate is the real number.
Read across 600 live markets on 31 July 2026, the distribution was 0.05 on
295 markets, 0.04 on 132, 0.07 on 82, and no rate at all on the remaining 91.
All 600 also carried takerOnly: true, which confirms the zero
maker fee rather than taking the documentation's word for it.
Two details of that live data will cost you, and neither is documented anywhere we could find:
- Where
feesEnabledis false, the rate field is absent, not zero. Code that reads the rate before checking the flag reads nothing, not a zero, and whether that becomes a crash or a silent default depends on the language. Check the flag first. takerBaseFeeandmakerBaseFeeare both 1000 on every market and are meaningless. They are defaults, not rates. Reading 1000 as basis points would give a 10% fee on everything. The rate lives infeeSchedule.rateand nowhere else.
Because the rate varies per market and over time, every fee this site stores is stamped with the schedule version it was priced under. Recomputing an old fee at today's rate rewrites history, and a record that does that is not a record.
Rounding, and where it actually bites
Polymarket's published schedule specifies no rounding step, so this page applies none. The fee is linear in stake, so there is no minimum-fee penalty on a small position: a $10 trade pays one tenth of what $100 pays at the same price. Price changes the share, size does not.
Kalshi is the contrast, and worth knowing because it is the other venue most
readers use. Its schedule reads roundup(0.07 × C × P ×
(1 - P)), rounding to the next cent, with the contract count
inside the ceiling. Rounding therefore applies to the whole trade
rather than per contract, and the consequence is counterintuitive: cheap
longshot contracts are where rounding eats the largest share of a small
position.
The formula makes that computable instead of anecdotal. Ten contracts at 5 cents produce a raw fee of 3.325 cents, which rounds up to 4, so you pay 20.3% more than the formula's own answer. The same ten contracts at 50 cents produce 17.5 cents, which rounds to 18, an uplift of 2.9%. A fixed one-cent quantum is a small tax on a large fee and a large one on a small fee, and the small fees live at the ends of the price range. With the break-even line above, the cheap contract is penalised twice.
Kalshi's own page carries the rest, including the maker rate our scanner assumes at 0.0175, which Kalshi's own two documents contradict each other on and the 0.035 half rate on S&P 500 and Nasdaq-100 tickers: Kalshi fee calculator.
A round trip pays twice, at two different prices
The calculator above prices one leg. Sell before resolution and you take the book again, paying a second taker fee priced off the exit price. That is not the same as doubling the entry break-even, and the difference usually runs in your favour.
A politics market at 4%. Buy 200 contracts at $0.50, a $100 stake, and the entry fee is $2.00. Sell those 200 at $0.90 and the exit leg prices at 90 cents, where the parabola is much lower: 200 × 0.04 × 0.90 × 0.10, which is $0.72. Round trip fees are $2.72, or 2.72% of stake, against the 4.00% you get by doubling the entry figure. A gross gain of $80.00 becomes $77.28 net.
Run it the other way and it flips: buying at 90 cents and selling at 50 costs $0.40 in and $1.11 out. The rule is short enough to keep in your head. Each leg is priced where it fills, and the leg nearer 50 cents is the expensive one. Holding to resolution pays the entry leg only, which is the cheapest way out of a taker position here.
How this page stays correct
Every coefficient here lives in one file, each carrying the source it came from and the date that source was read. The rates were verified against Polymarket's documentation on 29 July 2026 and confirmed against 600 live markets on 31 July 2026. The Kalshi comparison was verified on 30 July 2026 across three independent editions of their schedule, including the September 2022 CFTC rule filing, a primary regulatory document rather than a help page.
What happens next is a process, so it is described as one. The documentation page and a sample of live per-market rates will be re-read on a fixed schedule. If a coefficient moves, this page gets a dated correction with the superseded wording left visible above the new figure rather than replaced by it, and the stored schedule version increments so nothing priced at the old rate is silently re-priced at the new. No figures for that process are published until it has run.
Why bother: a competing calculator published a Polymarket taker coefficient of 0.0625 and ran it for five months. Polymarket does not use 0.0625. Every threshold that site displayed was wrong, and nobody caught it, because nothing on the page cited a source that could catch the drift. Of eleven competitor sites audited for this project, none has ever published a correction. That, not the presence of citations, is the gap.
Disclosure
This site is free to use, no sign-up and no paywall. The calculator runs in your browser and sends nothing anywhere.
One exception, stated here rather than buried: the Polymarket buttons on the scanner page carry a referral code, so Polymarket pays us if you open an account through one. It changes nothing about which candidates the scanner finds or how it ranks them. The Kalshi buttons pay nothing, and no link on this page pays anything at all.
Nothing on this page is financial advice, and prediction markets are not legal everywhere. Fees are only one cost of a trade: the spread you cross and the depth at your price will often matter more than the figures here.
Sources
- Polymarket trading fees docs.polymarket.com/trading/fees. Taker formula and the eleven category rates. Verified 29 July 2026.
- Polymarket Gamma API, per-market feeSchedule Read across 600 live markets on 31 July 2026. Observed rates 0.05 on 295, 0.04 on 132, 0.07 on 82, and absent on 91 where feesEnabled was false. takerOnly was true on all 600.
- Kalshi fee schedule kalshi.com/docs/kalshi-fee-schedule.pdf, cross-checked against the October 2025 edition and the September 2022 CFTC rule filing rule091222kexdcm003.pdf. Verified 30 July 2026. Used on this page only for the rounding comparison.
Fee coefficients are taken from each venue's published schedule and recorded with the date they were verified. Last updated .