Kalshi Parlays 2026: 170 Combos Priced
We priced 170 live Kalshi parlays against their own legs. The quote holds 10.68%, the fee adds 6.60%, and New Jersey sportsbooks hold 17.32% on parlays.
Affiliate disclosure: contains affiliate links; we may earn a commission at no extra cost to you. We have no commercial relationship with Kalshi. 18+ only — see BeGambleAware.org. Last updated 12 November 2026.
Key takeaways
- The quote is better than a sportsbook's. The fee hands the difference back. Across 170 live combos we priced on 12 November 2026, the quote alone implied a median hold of 10.68%. Kalshi's taker fee added a median 6.60% of stake on top, for an all-in 16.43%.
- New Jersey's sportsbooks held 17.32% on parlays over the same year. That is the regulator's own figure for the nine months to September 2026 — against 4.63% on everything that is not a parlay. All-in, a Kalshi combo lands 0.89 points under the parlay number.
- The fee bites because combos are longshots. As a share of stake it is 0.07 × (1 − P) — 6.3% at a 10¢ price, 3.5% at 50¢. Median combo ask on the day: 6.95¢.
- Almost none of the board is buyable. Of over 80,000 open cross-category combo markets, 73,866 were provisional shells and only 393 carried an offer you could hit.
- Combo maker fees are live on three series out of 16,264 —
quadratic_with_combo_maker_fees, all three combo series. No combo series carries a fee multiplier above 1.
On this page
Every guide to Kalshi parlays tells you the same thing: a market maker quotes your combo and builds margin into the price. Not one of them says how much. That is the only number that decides whether the product is worth using, and it is sitting in public data, so we went and measured it: 170 live combos, each priced against its own legs on the same afternoon, with Kalshi's fee arithmetic run on top of every one.
The answer has two halves pointing opposite ways, which is why nobody quotes it as one figure. The quote is materially cheaper than a sportsbook parlay — median implied hold 10.68% against the 17.32% New Jersey's regulator reports its books actually held on parlays this year. Then Kalshi's per-contract fee adds a median 6.60% of stake and the gap closes to 0.89 of a point. We ran the maths on every combo we could get a two-sided quote for.
What a Kalshi combo actually is
Kalshi calls them combos. The API calls them multivariate event collections, and its own field documentation is blunt about the mechanism: a collection defines a set of events, you pass two or more markets into a lookup call, and you get back one contract. Kalshi's functional_description for the cross-category collection reads: "The resulting market will only resolve to YES if every associated market resolves to YES."
That is a parlay in every respect that affects your money. One leg fails, the whole contract pays $0. All legs land, it pays $1. The arguments about whether it is legally a parlay — and three of the four pages ranking for this term spend real wordcount on them — change nothing about the arithmetic.
What is different is where the price comes from. A sportsbook multiplies its own already-margined odds, so a three-leg parlay compounds three helpings of vig. Kalshi's combos are quoted on request by a market maker, making the margin a single decision rather than a compounding one. That is the theory. Below is the practice.
The board is enormous and almost none of it is buyable
Start with scale, because the listing count is wildly misleading. Kalshi's collection endpoint returned 1,406 live collections across 16 series, carrying 21,200 associated-event entries between them. One of them — the cross-category collection, the one that lets you staple an NFL game to a tennis match — exposes 3,014 events on its own.
| Series | Collections | Min legs | Max legs | Median events offered |
|---|---|---|---|---|
| NBA single game | 1,113 | 2 | no cap | 8 |
| NFL single game | 223 | 2 | no cap | 10 |
| Cities weather | 35 | 2 | 7 | — |
| Mentions | 13 | 2 | no cap | 1 |
| NFL multi-game (extended) | 8 | 2 | no cap | 123 |
| Cross-category | 2 | 2 | no cap | 3,014 |
| All 16 series | 1,406 | 2 | no cap on 1,368 | 9 |
There is no leg limit on 1,368 of the 1,406 collections. This is the one hard fact the rest of the internet does not have: competitors say "no stated cap" and leave it there, but the cap is a field, and it is set to zero — meaning unbounded — nearly everywhere. The exceptions are small and specific: the 35 city-weather collections cap you at 7 legs, two Oscars collections at 10, a Grammys one at 4. Live markets bear that out. In the cross-category series we found contracts running from 2 legs up to 165.
Now the part that matters more. We paged the entire cross-category series and stopped at 80,000 open markets with the cursor still going:
| Cross-category combo markets, 12 Nov 2026 | Count | Share |
|---|---|---|
| Open markets paged (cursor not exhausted) | 80,000 | 100% |
Flagged is_provisional — a shell someone built and left | 73,866 | 92.3% |
| Has ever traded (volume above zero) | 6,134 | 7.7% |
| Buyable now — an offer below $1.00 with size behind it | 393 | 0.49% |
So the board is a graveyard of abandoned quote requests with a thin layer of live product on top. That is not a criticism — it is how a request-for-quote venue looks from outside, since every combo anyone ever asked about stays listed. But it means any count of "Kalshi parlay markets" is close to meaningless, and it is why the measurement below works with 170 contracts rather than 80,000.
How we priced 170 Kalshi parlays
The method the better guides describe but never run at scale: for an uncorrelated combo, fair value is the product of its legs' prices, and any excess is the house's margin.
We took a 12,000-market slice of the cross-category series, captured inside a minute so the quotes are comparable, and fetched every leg's order book straight afterwards. 304 had an offer we could hit; 277 of those had a two-sided quote on every leg, which is what a mid requires.
Then the filter that matters, and the one no competitor applies: 104 of the 277 drew two or more legs from the same event. Multiplying prices assumes independence, and a team to win stapled to that same team's total is nowhere near independent, so those 104 are excluded rather than quietly averaged in. Three more were dropped because their fair value fell below Kalshi's $0.0001 minimum price, where the tick floor alone forces a huge percentage premium. That leaves 170 independent-leg combos, median 7 legs, running from 2 to 30.
Two benchmarks, because they answer different questions. The product of leg mids gives the combo's total cost against fair value. The product of leg asks gives something sharper: whether the quote beats legging the same position in yourself, paying each spread.
The combo quote is cheaper than you would guess
Median premium over the product of leg mids: 11.96%. The interquartile range runs 4.98% to 30.00%. Against the product of leg asks it collapses to 3.57%, with the lower quartile actually negative at −0.22%.
Read those two numbers together, because the gap between them is the finding. Most of what a combo costs you is not combo margin at all — it is the legs' own spreads, compounded. Kalshi's quoter is adding roughly 3.6% for the service of bundling, and 47 of 170 combos were cheaper than assembling the same legs yourself at their asks.
| Legs | Combos | Median premium vs leg mids | Median all-in hold (with fee) |
|---|---|---|---|
| 3 | 18 | 4.74% | 9.77% |
| 4 | 14 | 5.37% | 10.75% |
| 5 | 11 | 6.52% | 11.49% |
| 6 | 18 | 10.99% | 15.12% |
| 7 | 19 | 8.66% | 14.07% |
| 8 | 20 | 12.45% | 17.02% |
| 9 | 9 | 11.54% | 16.42% |
| 10 | 9 | 24.82% | 25.28% |
| 13 | 7 | 14.81% | 18.71% |
| 16 | 5 | 31.56% | 30.07% |
The premium scales with leg count, and it scales faster than the quoter's own risk does. Three legs cost you under 5% over fair. By ten it is 24.82%, by sixteen 31.56%. Leg buckets with fewer than five observations are left out of that table entirely rather than presented as a trend.
One honest caveat on the tail. The mean premium across these 170 is useless — it is dragged past 200% by a handful of deep-longshot quotes, the worst a 12-leg contract asking $0.0005 against a fair value of $0.00000126. Those are real quotes, but they are quotes nobody is expected to take, so every figure here is a median or a quartile.
Then Kalshi's fee takes the advantage back
Kalshi's taker fee is 0.07 × contracts × P × (1 − P), rounded up. We checked all 17 Exotics and combo series and every one carries a fee multiplier of exactly 1 — the same coefficient as a plain NFL game market. There is no combo surcharge.
It is also where the product quietly loses its edge, and the reason is arithmetic rather than policy. Divide that fee by the price you pay and the price cancels: the fee as a share of your stake is 0.07 × (1 − P). The cheaper the contract, the bigger a bite the fee takes. Combos are longshots almost by construction — our median ask was 6.95¢, with a quartile down at 1.80¢ — which puts the fee at its most expensive end.
| Cost component (170 combos, median) | Lower quartile | Median | Upper quartile |
|---|---|---|---|
| Premium over product of leg mids | 4.98% | 11.96% | 30.00% |
| Premium over product of leg asks | −0.22% | 3.57% | 15.42% |
| Implied hold from the quote alone | 4.74% | 10.68% | 23.07% |
| Taker fee as a share of stake | 5.78% | 6.60% | 7.14% |
| All-in implied hold (quote + fee) | 10.17% | 16.43% | 27.36% |
Here is one real combo from the day, so you can see the whole stack on a single ticket. Three legs: Dallas to win, Tyrese Maxey 25+ points, Cam Skattebo 60+ rushing yards, at mids of 33.50¢, 55.00¢ and 48.50¢. Fair value multiplies out to 8.94¢. Kalshi's offer was 9.30¢ for 1,612 contracts — a 4.1% premium over fair, and 0.3% cheaper than buying the three legs at their asks. A good quote, in other words. The fee was 0.60¢ per contract: 6.5% of the stake, nearly twice the quoter's own margin, and it turns a 4.1% premium into a 9.7% all-in hold.
If you take one thing from this article, it is that on a Kalshi combo the fee is usually the larger cost, and it is the one nobody writes about. Our existing per-market fee breakdown has the same coefficient read across all 15,393 series for single contracts.
How Kalshi parlays compare to a real sportsbook parlay
Nobody in this niche quotes a sportsbook's parlay hold from a primary source, because operators do not break hold out by bet type. One regulator does it for them. New Jersey's Division of Gaming Enforcement reports sports wagering win and handle split by wager category every month, which makes parlay hold a published fact rather than an estimate.
| New Jersey, nine months to Sept 2026 | Handle | Win | Hold |
|---|---|---|---|
| Parlay | $2.687bn | $465.3m | 17.32% |
| Baseball | $1.064bn | $51.0m | 4.79% |
| Football | $627.7m | $28.9m | 4.61% |
| Basketball | $1.739bn | $42.2m | 2.42% |
| Everything that is not a parlay | $5.614bn | $260.2m | 4.63% |
| All sports wagering | $8.302bn | $725.5m | 8.74% |
Parlays were 32.4% of New Jersey's handle and 64.1% of its win. They hold 3.7 times what the rest of the book holds. That is the business Kalshi has walked into, and it explains why combos went from under 1% of its sports volume at launch to around a fifth of it by March 2026, on Legal Sports Report's compilation of the weekly data.
Set the two side by side. Kalshi's median all-in 16.43% against New Jersey's 17.32%: a real edge, but a narrow one, and it only exists because the quote is good enough to absorb a fee that is proportionally brutal. On the quote alone, 116 of 170 combos beat the New Jersey parlay number. Once the fee is in, 89 of 170 do — a coin flip.
One caveat on the comparison: New Jersey's figure is realised hold over nine months and billions of dollars, while ours is expected hold implied by quotes on one afternoon. What survives that caveat is the shape. Take tight three- and four-leg combos and you pay around 10% all-in, which genuinely halves the parlay cost. Take ten legs and you are at 25%, which is worse than the sportsbook.
Price the legs before you accept the quote
Every combo quote can be checked in thirty seconds against the markets it is built from. Polymarket runs the same multi-leg maths on a different fee model, and comparing the two on one ticket is the cheapest edge in this niche.
Visit Polymarket.com →Affiliate link to Polymarket; we have no commercial relationship with Kalshi. 18+ only. Polymarket restricts access from the United States, Canada and other listed jurisdictions. Prediction markets carry a substantial risk of loss. Nothing here is financial advice.
Combo maker fees went live on three series
There has been reporting that Kalshi was preparing to charge makers on parlays, where previously the people quoting combos paid nothing. Kalshi's own series records now settle it. Each series carries a fee_type, and across all 16,264 series live on 12 November 2026 there are exactly three values:
| Fee type | Series | What it means |
|---|---|---|
quadratic | 16,101 | Taker fee only — resting orders are free |
quadratic_with_maker_fees | 160 | Makers charged too (the headline sports series sit here) |
quadratic_with_combo_maker_fees | 3 | A combo-specific maker charge |
All three are combo series: the two cross-category series and the multi-game sports one. Every other combo series — including all 1,113 NBA single-game collections and all 223 NFL ones — is still plain taker-only. The change is real, it is live, and it has been rolled out to the two busiest combo series and nowhere else. Whether that reaches the rest is a thing to re-read in a month; if it does, expect quoters to widen, which shows up as a higher premium in the table above long before it shows up in any announcement. Our maker fee census covers the single-contract side of the same question.
Where the quote is wrong in your favour
A quoter pricing thousands of bespoke combinations will misprice some, and 12 of our 170 were offered below the product of their legs' mids — the best at −9.73%. One three-leg soccer combo was offered at 21.10¢ against a fair value of 22.53¢, which is still a negative hold after the fee.
Do not read that as free money, for three measurable reasons. Independence is an assumption, and a combo that looks underpriced may hold legs correlated in a way multiplication misses. The size is small — that soccer combo had 947 contracts behind it. And a mid is not a tradeable price, which is why the ask-based benchmark is the stricter test; on that basis 47 of 170 still came out ahead.
The practical version: the combo builder is worth using as a price check even when you have no intention of buying a combo. If the quote comes back below the product of the legs you were going to back one by one, the combo is the cheaper way to express the same view — and if it comes back 30% above, you have learned that in thirty seconds. That is the same discipline we applied to multi-outcome fields on Polymarket, where the median overround was 4.15% and five of 496 fields were genuinely risk-free.
What to check before you buy a combo
- Multiply the legs before you accept. Mids, not asks, for a fair-value read. More than about 12% over it and you are paying above the typical combo.
- Add 0.07 × (1 − price) to the cost. At a 5¢ combo that is 6.65% of your stake before the quoter's margin.
- Keep the leg count down. Three and four legs cost about 10% all-in; ten legs cost 25%.
- Treat same-event legs as unpriceable from outside. Multiplication does not work on correlated legs, so you cannot tell a fair quote from a bad one.
- Check the size, not just the price. A good quote on 7 contracts is not a trade.
- Check whether the series charges combo maker fees if you intend to rest an order rather than take one.
And the structural point, whichever venue you use: parlays hold 3.7 times what straight bets hold in New Jersey, and Kalshi's version is cheaper by roughly a point, not by a category. The combo is a better-priced version of an expensive product. Our liquidity comparison and payout timing data cover what happens once the ticket is on.
Related guides
- Kalshi Fees by Market 2026: 195 Exceptions — the same fee coefficient read across every single-contract series.
- Kalshi Maker Fees 2026: 163 Series Billed — who pays for resting an order, and where.
- Polymarket Arbitrage 2026: 716 Fields Priced — the same multiplication applied to multi-outcome fields.
- Kalshi vs Polymarket Liquidity 2026 — whether the size behind a quote is real.
- Kalshi vs Polymarket Odds 2026: 626 Prices — which venue prices the same event better.
- Kalshi Payout Time 2026: 15,382 Timed — how long you wait for a winning leg to settle.
Kalshi parlays FAQ
Can you parlay on Kalshi?
Yes, through the combo builder, which Kalshi's API calls a multivariate event collection. You pass two or more eligible markets in and get one contract back that pays $1 only if every leg resolves YES, and $0 otherwise. On 12 November 2026 there were 1,406 live collections across 16 series to build from.
How many legs can a Kalshi parlay have?
The minimum is two and on 1,368 of 1,406 collections there is no maximum at all — the cap field is set to unbounded. The exceptions are the 35 city-weather collections, capped at seven legs, two Oscars collections at ten and a Grammys collection at four. In the live cross-category series we found contracts carrying anything from 2 to 165 legs.
Are Kalshi combos cheaper than a sportsbook parlay?
On the quote, clearly. Across 170 combos the quote alone implied a 10.68% median hold against the 17.32% New Jersey's regulator reports its sportsbooks held on parlays over the nine months to September 2026. Once Kalshi's taker fee is added the all-in figure is 16.43%, which is only 0.89 points better. Short combos keep most of the advantage; long ones lose it entirely.
What fee does Kalshi charge on a combo?
The same quadratic taker fee as any other contract: 0.07 x contracts x P x (1 - P), rounded up, with a fee multiplier of 1 on all 17 combo and Exotics series. There is no combo surcharge. But as a share of your stake the fee works out at 0.07 x (1 - P), so it is worst on cheap contracts: 6.3% of stake at a 10 cent price against 3.5% at 50 cents. Our measured median was 6.60%.
How do I tell whether a combo quote is fair?
Multiply the legs' mid prices together. For legs in different events that product is fair value, and the typical quote we measured sat 11.96% above it, with the middle half of combos between 4.98% and 30.00%. The stricter test is the product of the legs' ask prices: if the combo is below that, it beats buying the legs one at a time, which 47 of our 170 did.
Can I sell a Kalshi combo before it resolves?
In principle yes — a combo is a tradeable contract with its own order book, not a settled ticket. In practice liquidity is thin: of over 80,000 open cross-category combo markets, only 6,134 had ever traded and 393 carried a live offer with size. Assume you may not get a fair exit price.
Why are there so many Kalshi combo markets with no price?
Because combos are quoted on request, and every combination anyone ever asked about stays listed. 73,866 of the 80,000 cross-category markets we paged were flagged provisional — shells built by a request, then abandoned. Any headline count of Kalshi parlay markets is counting those.
Method and sources
Measured on 12 November 2026. Structural figures are a complete read of Kalshi's public multivariate event collections endpoint — all 1,406 collections, with size_min, size_max and the associated-event lists read verbatim and field meanings taken from Kalshi's documentation rather than inferred. Fee types come from a full read of all 16,264 series. One field we deliberately do not report: active_quoters is present but empty on all 20,940 event entries without authentication, so it is no evidence either way.
The price measurement is a dated cohort, not a census. It uses the first 12,000 open markets the cross-category cursor returned, captured within about a minute so the combo quotes and leg books are comparable, which is also why we did not price the full 80,000 — a sweep that long would compare quotes taken minutes apart. From that slice: 304 buyable, 277 with every leg two-sided, 104 excluded for drawing two legs from one event, 3 excluded for a fair value below the $0.0001 minimum price, leaving 170. Fees are computed from Kalshi's published formula with the multiplier of 1 we read off each series, rounded the way Kalshi rounds an order. Implied hold is the overpay as a share of total outlay, so it is comparable to a sportsbook's hold on handle.
Sportsbook figures are the New Jersey Division of Gaming Enforcement's September 2026 release, year-to-date table, which the agency marks unaudited and subject to adjustment; the non-parlay row is our subtraction of the parlay row from the total, and the hold percentages are our division of win by handle. Kalshi's parlay volume trend is Legal Sports Report's compilation, running to March 2026 and cited as theirs. Kalshi is a CFTC-designated contract market, which is why its combos are financial contracts rather than wagers — a distinction with no effect on the cost arithmetic here.
Quotes move constantly and the premium figures will not reproduce exactly on another day. The structure — leg caps, fee formula, fee types, the tradeable share of the board — moves slowly, and those are the figures to rely on.