US Open 2026 Men's Odds: The AI Model Verdict Vs. The Kalshi Market
Kalshi's US Open men's board has already made up its mind. Traders have installed Jannik Sinner above 50 cents, twice the price of Carlos Alcaraz, in a 33-player field that has traded roughly 655,000 contracts. We asked eight AI models to price the entire field without showing any of them a single market number, then let them read each other's anonymized reasoning and revise. They came back with a different shape at the top: a near coin flip between two players the market has split two-to-one.
The Quick Answer
Kalshi prices Jannik Sinner at 51.5 cents to win the 2026 US Open. Our eight-model, price-blind panel blends him to just 29.2%, the single biggest fade on this board, with defending champion Carlos Alcaraz almost level at 26.0% against a 24.5-cent price. Below the duopoly, the panel doubles Novak Djokovic (6.0% vs 3 cents) and Daniil Medvedev (4.0% vs 2 cents). The full board, every seat's number, and the argument that moved three models in the revision round are all below.
The Market At A Glance
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| The Event | "US Open Men's Singles Winner" (Kalshi event KXATP-26USO), 33 mutually exclusive player contracts |
| The Contract | Per the settlement rules, verbatim: "If Jannik Sinner wins the 2026 US Open Men's Singles professional tennis tournament, then the market resolves to Yes" (identical wording per player) |
| Tournament Dates | Main draw August 30 through September 13, 2026, New York, hard court; contracts carry an expected expiration of September 14, 2026 (05:00 UTC) |
| Prices Below As Of | August 1, 2026, 12:31 UTC (live bid/ask midpoints) |
These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate.
When Is The 2026 US Open?
The 2026 US Open main draw starts Sunday, August 30, under the Sunday-start format the tournament adopted in 2025, and runs through the finals weekend of September 13 in New York. That gap is exactly why this board is interesting now: the draw has not been made, the summer hard-court tune-ups are still being played, and four weeks of dead air separate today's prices from the first ball. Prices set this far out are set on reputation, and reputation is precisely where our panel and the market disagree.
Want the re-scored board when the draw drops? The panel will re-run this market once the bracket is out and the tune-up results are in the books. The email signup at the end of this page gets that update first, wherever you are reading from.
The vehicle is The Weekly PM Market Brief: the 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next — free, one email, Sundays.
What Actually Wins The US Open
Before any name, the shape of the event. Three structural facts did most of the work in the panel's reasoning, and all three come from data fetched the morning these verdicts were generated:
- The 2026 season has three different major champions. Carlos Alcaraz won the Australian Open over Novak Djokovic in January, Alexander Zverev broke through for his first major at the French Open, and Jannik Sinner defended his Wimbledon title over Zverev in July. One player winning everything is not this year's story.
- Recent US Open history is a short list. The last five champions in New York: Alcaraz (2025, over Sinner in the final), Sinner (2024, over Taylor Fritz), Djokovic (2023, over Medvedev), Alcaraz (2022), Medvedev (2021, over Djokovic). All four men on that list are on this board, and all four sit in the panel's top five.
- Rankings are a form proxy with a lag. Sinner leads the ATP rankings with 13,450 points to Zverev's 8,480 and Alcaraz's 8,160 (as of July 27, 2026). The panel treated that 5,000-point cushion as a season-long consistency signal, and argued about how much it should count against Alcaraz holding the two freshest hard-court major results.
Best-of-five durability over seven rounds is the filter that turns a 128-player draw into the short list above. The market and the panel agree on the list. They disagree on how the top of it splits.
The 2026 US Open Men's Board
First the phone-friendly version: market price against the panel's blended number, sorted by the blend. The blend is the equal-weight mean of all eight models' final numbers after a revision round where each model read the others' anonymized reasoning.
| Player | Market (mid) | AI Blend | Gap |
|---|---|---|---|
| Jannik Sinner | 51.5¢ | 29.2% | −22.3 |
| Carlos Alcaraz | 24.5¢ | 26.0% | +1.5 |
| Alexander Zverev | 8.5¢ | 10.7% | +2.2 |
| Novak Djokovic | 3¢ | 6.0% | +3.0 |
| Daniil Medvedev | 2¢ | 4.0% | +2.0 |
| Any other listed player (13 names) | ~6.5¢ combined | 3.7% | −2.8 |
| Ben Shelton | 2¢ | 3.0% | +1.0 |
| Taylor Fritz | 2.5¢ | 2.8% | +0.3 |
| Felix Auger-Aliassime | 1.5¢ | 2.6% | +1.1 |
| Any player not listed at all | n/a | 2.2% | n/a |
These are model estimates, not predictions of fact and not financial advice. The gap column is the blend percentage minus the raw market price in cents; it mixes a probability with an un-adjusted price from an order book with spread, so read it as direction and size, not a precise edge. The 33 raw mids sum to about 111 cents rather than 100 because each contract trades in its own book.
Every Seat's Number
The full per-model board for all 20 individually priced players, plus the two buckets each model was required to price so its distribution sums to 100. Values are percentages from each model's final (post-revision) board.
| Player | Market | Fable | Opus | Sonnet | GPT | Gemini | GLM | Kimi | DeepSeek | Blend |
|---|---|---|---|---|---|---|---|---|---|---|
| Jannik Sinner | 51.5¢ | 30.0 | 28.0 | 28.0 | 28.7 | 30.5 | 30.0 | 31.5 | 26.5 | 29.2 |
| Carlos Alcaraz | 24.5¢ | 28.0 | 25.5 | 22.7 | 24.8 | 28.0 | 25.5 | 27.0 | 26.5 | 26.0 |
| Alexander Zverev | 8.5¢ | 10.3 | 10.0 | 11.0 | 11.2 | 10.5 | 11.0 | 10.5 | 11.0 | 10.7 |
| Novak Djokovic | 3¢ | 5.5 | 5.5 | 7.0 | 6.5 | 6.0 | 6.0 | 5.0 | 6.5 | 6.0 |
| Daniil Medvedev | 2¢ | 3.0 | 3.5 | 4.0 | 4.6 | 4.5 | 4.5 | 3.5 | 4.0 | 4.0 |
| Ben Shelton | 2¢ | 2.8 | 3.5 | 3.0 | 3.0 | 3.0 | 3.0 | 2.5 | 3.0 | 3.0 |
| Taylor Fritz | 2.5¢ | 2.4 | 2.9 | 2.8 | 3.0 | 3.0 | 2.5 | 2.5 | 3.0 | 2.8 |
| Felix Auger-Aliassime | 1.5¢ | 2.6 | 3.0 | 2.5 | 2.8 | 2.5 | 2.5 | 2.5 | 2.5 | 2.6 |
| Alex de Minaur | 0.5¢ | 1.5 | 1.5 | 2.0 | 2.0 | 2.0 | 1.5 | 1.2 | 2.0 | 1.7 |
| Flavio Cobolli | 0.5¢ | 1.2 | 1.5 | 1.5 | 1.6 | 2.0 | 1.5 | 1.5 | 1.5 | 1.5 |
| Jack Draper | 0.5¢ | 1.2 | 2.0 | 1.0 | 1.3 | 0.8 | 1.5 | 0.8 | 1.0 | 1.2 |
| Frances Tiafoe | 0.5¢ | 0.6 | 0.9 | 1.2 | 0.7 | 0.8 | 1.0 | 0.8 | 1.0 | 0.9 |
| Jakub Mensik | 1.5¢ | 0.8 | 1.1 | 0.6 | 0.9 | 0.2 | 1.5 | 0.8 | 0.5 | 0.8 |
| Lorenzo Musetti | 0.5¢ | 0.7 | 1.0 | 0.6 | 0.8 | 0.5 | 0.5 | 0.8 | 0.8 | 0.7 |
| Tommy Paul | 0.5¢ | 0.6 | 0.7 | 0.5 | 0.6 | 0.5 | 0.5 | 0.4 | 1.5 | 0.7 |
| Learner Tien | 0.5¢ | 0.9 | 0.9 | 0.8 | 0.4 | 0.1 | 0.5 | 0.8 | 0.5 | 0.6 |
| Alexander Bublik | 0.5¢ | 0.8 | 0.9 | 0.8 | 0.6 | 0.2 | 0.5 | 0.8 | 0.3 | 0.6 |
| Joao Fonseca | 1.5¢ | 1.0 | 1.2 | 0.4 | 0.4 | 0.1 | 0.5 | 0.5 | 0.5 | 0.6 |
| Arthur Fils | 0.5¢ | 0.5 | 0.7 | 0.4 | 0.5 | 0.2 | 0.3 | 0.4 | 0.6 | 0.5 |
| Rafael Jodar | 1.5¢ | 0.4 | 0.3 | 0.2 | 0.1 | 0.1 | 0.3 | 0.2 | 0.2 | 0.2 |
| Other listed (13 names) | ~6.5¢ | 3.2 | 3.3 | 6.0 | 3.5 | 3.0 | 3.0 | 3.5 | 3.8 | 3.7 |
| Field (not listed) | n/a | 2.0 | 2.0 | 3.0 | 2.0 | 1.5 | 2.0 | 2.5 | 2.8 | 2.2 |
Every seat on this panel is graded against real market settlements — records to date: Claude Fable 85% on 331 graded calls · Claude Opus 86% on 382 graded calls · Claude Sonnet 84% on 363 graded calls · GPT 83% on 1,693 graded calls · Gemini 85% on 1,718 graded calls · GLM 80% on 1,874 graded calls · Kimi 84% on 1,863 graded calls · DeepSeek 80% on 1,877 graded calls. Recomputed daily; the full scoreboard is public.
These are model estimates, not predictions of fact and not financial advice. A note on the panel. All eight seats answered both rounds: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT, Gemini Pro, GLM, Kimi, and DeepSeek. Each priced the whole field in one pass, price-blind, from a data card fetched the same morning (the season's three major results, the last five US Open champions, the live ATP top 20, and a verified line on every listed player), with the 13 lowest-activity names priced as one combined bucket. Every number lands in our backtest ledger and gets graded against the real settlement on our public Model Verdict scoreboard when the champion is decided.
More live boards from the same panel: National Bank Open winner, the last hard-court Masters before this draw, where Ben Shelton and Rafael Jodar are level at the top at 22.5¢ and 22¢ · US Open women's champion, where Aryna Sabalenka sits at 28¢ · Ballon d'Or 2026, where Harry Kane leads at 58¢. Prices fetched August 9, 2026.
Jannik Sinner: The 51-Cent Favorite The Panel Fades
The whole board turns on one question: is the world No. 1 a true 50/50 proposition against the field? The market says yes at 51.5 cents. All eight models say no, and none of them printed higher than 31.5.
The case for Sinner is real and the panel priced it first: he is No. 1 by nearly 5,000 points, he just defended his Wimbledon title over Zverev, he won this tournament in 2024, and he made the final here last year. Kimi, the panel's most aggressive Sinner backer, still cut its own number from 37 to 31.5 in the revision round, explaining: "Alcaraz owns the two freshest hard-court major results (2025 US Open final over Sinner, 2026 Australian Open), while Sinner's ranking lead is accumulated across all surfaces, so a 10.5-point gap was unjustified. I still keep Sinner first on the 5,300-point lead plus his 2024 title." — Kimi
That is the panel's position in miniature. Sinner deserves the top line. He does not deserve twice Alcaraz's price, because a seven-round, best-of-five tournament with a draw that does not exist yet carries too much variance for any player to be a coin flip against the field. A 51.5-cent price re-scales to roughly 46% of this board's total; the blend pays 29.2%. That 17-point gap is the widest disagreement on the board and one of the widest we have published on any current market.
Carlos Alcaraz: The Defending Champion The Panel Holds Level
Alcaraz at 24.5 cents is the number the panel thinks the market got most wrong in relative terms. He is the defending champion in New York, a two-time US Open winner, and he beat Sinner in last year's final on this exact court. In January he won the Australian Open over Djokovic, completing the career Grand Slam as the youngest man to do it. Those are the two most recent hard-court majors played, and he won both of them.
The counterweight is the rankings table: Alcaraz sits third at 8,160 points, behind even Zverev, and the panel took that seriously as evidence his season outside the majors has been thinner than Sinner's. That argument is what moved Claude Fable, the one seat that had Alcaraz first in round one, to flip its top two in the revision round: "Forecaster C's point that Alcaraz ranks only No. 3 while holding AO-champion and USO-defending points — implying thin 2026 results everywhere else, versus Sinner's ~5,300-point cushion — is a card fact I had underweighted as a consistency signal for seven rounds of best-of-five." — Claude Fable (the forecasters read each other anonymized, as lettered seats)
The blend lands at 26.0% against a 24.5-cent price: agreement on the level, disagreement on the spread. The market prices Sinner as a 2-to-1 favorite over Alcaraz. The panel says the true gap between them is about three points.
Alexander Zverev: The New World No. 2 Without A Hard-Court Major
Zverev's 2026 resume is the best of his career: French Open champion, Wimbledon finalist, world No. 2, and now a finalist at all four majors. The market prices that at 8.5 cents; the panel blends him to 10.7%, with every single seat between 10.0 and 11.2, the tightest agreement on the board.
The ceiling argument was just as unanimous. Both of Zverev's 2026 major finals ended against the top two (he beat Flavio Cobolli for the French title, then lost to Sinner at Wimbledon), and the panel kept coming back to the same structural point: his breakthrough came on clay, and he has never won a hard-court major. GLM put the tier boundary plainly: his best-of-five pedigree "is finally converting," but the US Open champion pattern in the data runs through Sinner, Alcaraz, Djokovic, and Medvedev, not through him. A 10-11% number says: clearly third, clearly alive, clearly below the duopoly.
Novak Djokovic And Daniil Medvedev: The Pedigree Doubles
The two most interesting small numbers on the board belong to former champions. Djokovic, at 39, made the 2026 Australian Open final in January and won this event as recently as 2023; the market has him at 3 cents and the panel doubles that to 6.0%. Medvedev, the 2021 champion and 2023 finalist, ranked No. 7, trades at 2 cents against a 4.0% blend.
Both numbers were argued in the revision round rather than assumed. GLM trimmed Medvedev after peers pointed at his silent 2026: "the card shows zero 2026 major results for him — his US Open pedigree is real but current form is silent." — GLM. Claude Opus moved the other way and raised him, on the grounds that Medvedev's 2021 title and 2023 final are "the only New York pedigree outside the Sinner/Alcaraz/Djokovic cluster," and that absence of 2026 results in the data is not evidence of decline. The blend settling at exactly double the market price, on both names, is the panel saying: cheap pedigree at a hard-court major is systematically underpriced at the 2-to-3-cent level.
Dark Horses The Panel Won't Dismiss
Every seat was required to name sub-5% players it would least want to dismiss, with a stated structural reason. Four names kept recurring:
- Ben Shelton (2¢, 3.0% blend). The No. 8 player in the world at his home major. DeepSeek's case: "his explosive serve and lefty power thrive on New York hard courts." — DeepSeek
- Flavio Cobolli (0.5¢, 1.5% blend). A 2026 French Open finalist and top-10 debutant. Claude Fable's case: "proven best-of-five stamina in a five-set major final." — Claude Fable
- Taylor Fritz (2.5¢, 2.8% blend). The 2024 US Open runner-up, still ranked No. 10, with the most recent American final here.
- Learner Tien and Joao Fonseca (0.5¢ and 1.5¢). The two youngest names the seats kept flagging: Tien, born in 2005, is already No. 16 and, in Claude Fable's words, "an American peaking into his home slam"; Fonseca, born in 2006, is what Claude Opus called "the kind of fast-rising power baseliner whose ceiling the rankings lag badly." Fable put Jakub Mensik in the same bucket: "the exact profile that outperforms lagged rankings at hard-court majors."
The Traps: Big Gaps That Are Not Value
Honesty about the table's edges matters more than a clean story:
Rafael Jodar And Jakub Mensik: The 1.5-Cent Names The Panel Fades
Three names trade at 1.5 cents (Jodar, Mensik, Fonseca) and the panel blends all three at or below 0.8%. Jodar is the sharpest case: a Spaniard born in 2006, at a career-high No. 23, but the panel's blend of 0.2% makes his 1.5-cent price the board's biggest relative overprice. At these levels, though, one cent is close to the exchange's minimum tick, spreads are wide, and the "edge" in fading a 1.5-cent contract is mostly theoretical. Both men have since reached the quarterfinals in Montreal, where Jodar is priced as a joint favorite; the same panel scored that eight-man field in the National Bank Open model verdict.
The Missing No. 12
The sharpest catch of the revision round came from Claude Opus, checking the board list against the rankings table: "Jiří Lehečka, world No. 12 at 2,470 points, is not on the board at all." — Claude Opus. A champion who is not listed resolves every contract on this board to No. Two seats raised their field bucket specifically on that observation, and the blend's 2.2% on "any player not listed" is not a rounding artifact; it is a real, structural number a 33-name board cannot cover.
The 0.5-Cent Floor Is Not A Probability
From de Minaur down, most contracts trade at a 0.5-cent mid on thin volume. Those prices cannot go much lower, so a model blend of 1.5% against 0.5 cents is not a three-fold edge you can actually collect; it is a statement about which longshots the panel thinks are mispriced within the noise. Treat everything below 1% blend as texture, not signal.
How To Read These Prices
The models can tell you where they part from the crowd; the number in front of you still has to be read in context.
Work One Example: Sinner At 51.5 Cents
A 51.5-cent YES quote is not a clean "51.5% chance." Each of the 33 contracts trades in its own order book with a bid/ask spread, so the mids across this mutually exclusive field sum to about 111 cents rather than 100. Re-scale each price against that total and Sinner's 51.5 cents works out to roughly 46% of the board; Alcaraz's 24.5 cents becomes about 22%. Set those against the panel's 29.2% and 26.0% and the real shape of the disagreement appears: the panel is not saying Sinner will lose, it is saying the market has priced a duopoly as a monarchy. That re-scaling habit is the same discipline as converting Kalshi's cents into real odds before comparing anything to a sportsbook line, and the mechanics of trading a board like this live in our guide to how to bet sports on Kalshi.
A few honest frictions before acting on any number above:
- Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and in an eligible state, and availability varies by state and can change.
- Liquidity And Spread. Below the top four names, most of this board trades on thin books at wide relative spreads (several contracts show a 0-to-1-cent quote). Getting filled at the printed mid is not a given, and prices recovered from thin books deserve extra skepticism.
- Time And Capital. These contracts are expected to expire September 14. Money committed today is locked through six weeks of tune-ups, a draw release, and two weeks of tennis.
- Not Advice. Everything here is analysis for information. These are model estimates, not predictions of fact and not financial advice.
For the day-to-day version of this work, our Kalshi picks hub tracks the boards we have scored across every category, our live odds screen does the price-comparison job on the sportsbook side, and our free expert picks cover the daily slate. The method itself has siblings: the same eight-seat, price-blind protocol scored the US Open women's board in July and the 2026 F1 championship field on live standings.
FAQ
When is the 2026 US Open men's singles? The main draw runs August 30 through September 13, 2026 on the hard courts in New York, with the men's final on the second Sunday. Each Kalshi contract resolves once a champion is decided; the exchange lists an expected expiration of September 14, 2026 (05:00 UTC).
Can I trade the Kalshi US Open men's winner market? Kalshi is a CFTC-regulated event-contract exchange open to users who are 18 or older and located in an eligible state, and availability varies by state as of August 2026. Nothing here is financial advice.
The Bottom Line
The market and the panel are reading the same season and pricing different tournaments. Kalshi is paying for the ranking table, where Sinner's 5,000-point lead looks like dominance. The panel is paying for the last two hard-court majors, both of which belong to Alcaraz, and for the variance of a 128-player, best-of-five fortnight whose draw does not exist yet. The result is the board above: Sinner still first, but at 29% rather than 51 cents, with the defending champion three points behind rather than 27. If the tune-ups or the draw change that picture, this page gets re-scored on the same URL. In the meantime, every number here sits in the backtest ledger, waiting to be graded against whatever happens on Arthur Ashe in September.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $166M traded
- Rocket Classic Winner · $85M traded
- 2027 Pro Football Champion · $52M traded
- Pro Baseball Champion · $41M traded
- Bitcoin price at the end of 2026 · $30M traded
- 2027 Pro Basketball Champion · $19M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub.
Verdicts generated August 1, 2026, after a revision round in which each model read the others' anonymized reasoning. Prices as of August 1, 2026, 12:31 UTC. Panel: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT, Gemini Pro, GLM, Kimi, and DeepSeek, all price-blind. These are model estimates, not predictions of fact and not financial advice.
Prices on this page are Kalshi's book. If you also trade on Polymarket, code OS4 gets new users a $20 bonus on a $10 deposit — affiliate link; terms as stated by Polymarket; 18+, availability varies by state.


