By Jake Hari. Prices are last trades as of September 10, 2026. Panel verdicts were generated August 22, price-blind, from the 2028 Senate map, the locked 2024 carryover, presidential-year base rates back to 1980 and the 2026 midterm environment.
Kalshi traders make Democrats the favorite to control the Senate after the 2028 election, with the last trade at 57 cents on the dollar. Eight AI models that priced the same map without seeing the market call it a coin flip, with a median of 51.
The market's read is almost identical to its price on a Democrat winning the White House on the same election night, which last traded at 58 cents. The panel says those are two different bets, and the reason starts with which seats are actually on the 2028 ballot.
Which Seats Are Actually On The 2028 Ballot?
Two thirds of this market is settled before a single 2028 ballot is printed. Of the 100 senators seated in January 2029:
- 33 Seats Are Already Locked. They were filled in 2024 and are not up again until 2030. That group splits 19 to 14 for Democrats, counting the two independents who caucus with them.
- 33 Seats Are Decided This November 3, in the 2026 midterms, which we score seat by seat on the 2026 balance-of-power board.
- 34 Seats Are On The Ballot In November 2028. This is the group the models spend their reasoning on.
That leaves 67 seats still to be decided across the two elections. Democrats need 32 of them for a majority. Republicans need 37.
A tied Senate is run by the vice president's party, so whichever party wins the White House in 2028 can hold the chamber with 50 seats. That is the only mechanical thread tying this market to the presidential race, and it is worth exactly one seat. The bigger link is correlation, and the panel splits on which way it cuts.
Two of the 34 seats have appointed senators in them: Marco Rubio left Florida's to become Secretary of State, and JD Vance left Ohio's to become vice president. Both get special elections on November 3, 2026 for the rest of the current term, then return to the regular cycle in 2028.
Why Won't The Panel Follow The Price?
Most of the models moved toward Republicans after reading each other, and nearly all of them moved for the same reason. The 2028 class is usually described as 19 Republican-held seats and 15 Democratic-held, counting Florida and Ohio, which sounds like Republicans have more to defend. Count the seats whose incumbent won with less than 54% of the vote in 2022 and five of them sit on the Democratic side: Nevada, Pennsylvania, Arizona, Georgia and New Hampshire. Only Wisconsin and North Carolina are in that range for Republicans, though each of those two was individually tighter than every Democratic seat but Nevada.
Utah and Alaska look close on paper too, but Mike Lee's opponent was an independent with no Democrat on the ballot, and Lisa Murkowski beat a fellow Republican in a ranked-choice count. A low winning share in a safe Republican state is no opening for Democrats, which is why the panel counted two live Republican targets rather than four. Opus, the most Republican-leaning model on the board, put the exposure in one line:
The decisive fact is exposure reversal: Democrats defend five sub-54% seats (NV 48.8, PA 51.2, AZ 51.4, GA 51.4, NH 53.5) against only two marginal R seats (WI 50.4, NC 50.5). — Opus
Sonnet gave Democrats a modest gain this November and still found that Republicans reach a majority by simply holding every seat they already have in the 2028 group:
Republicans only need Class 3 to hold serve (net zero) to reach 51; Democrats need a net +2 swing, and the marginal-seat map doesn't clearly hand them that. — Sonnet
The seat almost every model named first is Nevada, where Catherine Cortez Masto's 2022 win was the narrowest of anyone on the 2028 ballot, with John Fetterman, Mark Kelly and Raphael Warnock the next-closest Democrats. Those four states are also presidential battlegrounds, which is exactly why the models split. Kimi described the fork directly:
Nationalization cuts both ways: a Democratic presidential win protects those very swing-state seats and exposes WI/NC; a Republican win flips NV and likely more. — Kimi
That explains why the two contracts move together, but it does not explain why they cost the same. Opus put a Democratic Senate at 42.5 against the market's 58-cent price on a Democratic presidency, because winning the White House still leaves the party holding every thin seat and then flipping one more.
Presidential years also rarely move the Senate much; the median swing since 1980 is 3 seats. Only two elections broke that pattern: Reagan's landslide, which flipped 12 seats, and 2008, when a financial crisis broke weeks before the vote. No model could name a comparable condition two years out, so none of them priced one.
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
The Board
| Outcome | Last trade | Book (bid / ask) | AI panel (median) | Gap |
|---|---|---|---|---|
| Democratic Majority | 57¢ · Sep 3 | 51¢ / 56¢ | 51.0% | Market +6.0 |
| Republican Majority | 46¢ · Aug 24 | 43¢ / 49¢ | 49.0% | Panel +3.0 |
Prices are the most recent trades as of September 10, 2026; the panel figure is the median of the eight models, so one outlier cannot drag it. The only trading since August 24, when 419 contracts went through on the Republican side at 46 cents, has been about 44 contracts on the Democratic side at 57 cents on September 2 and 3, one cent above the previous print, and that is the whole reason the top line moved. The book is thin: the tight quotes on both sides were pulled on August 27, restored the next day at the same prices without a single contract trading, then pulled again overnight into September 10 and restored that same afternoon. As of this update the Democratic side is quoted 51 cents bid against 56 asked and the Republican side 43 against 49, a cent tighter than before the pull, and still not one contract has traded on either side since September 3. A screen midpoint that swings 12 points and back inside a day on zero trades says nothing about the odds; it says an order was cancelled and re-entered. The better cross-check is Kalshi's combined market on the presidency, House and Senate together, which has traded about $380,000; add up its four Democratic-Senate scenarios and they last traded at roughly 56 cents, so the thin board is not out of line with the deeper one.
Every Model's Number
Each model priced the market without seeing it. Round 1 is its first independent answer. Round 2 is after reading the other seven's reasoning, anonymized, with the freedom to revise or hold.
| Model | Round 1 (D) | Round 2 (D) | Moved |
|---|---|---|---|
| DeepSeek | 42.5% | 63.0% | +20.5 |
| Gemini | 65.0% | 54.0% | −11.0 |
| ChatGPT | 56.0% | 53.0% | −3.0 |
| GLM | 48.0% | 53.0% | +5.0 |
| Fable | 57.0% | 49.0% | −8.0 |
| Sonnet | 57.0% | 46.0% | −11.0 |
| Kimi | 55.0% | 44.0% | −11.0 |
| Opus | 44.0% | 42.5% | −1.5 |
| Panel Median | 55.5% | 51.0% | −4.5 |
Every model on this panel is graded against real market settlements — records to date: Gemini 86% on 4,423 graded calls · GLM 81% on 2,437 graded calls · Kimi 82% on 2,414 graded calls · DeepSeek 80% on 2,462 graded calls. Recomputed daily; the full scoreboard is public.
- Every number here gets graded in public once the market settles, on the full graded scoreboard.
More live boards from the same panel: 2026 Senate control, Republicans 53¢ · 2026 House control, Democrats 86¢ · 2028 presidential race, Democrats 58¢. Prices fetched September 10, 2026.
Where The Panel Changed Its Mind
The revision round moved this board more than most, and it moved away from the price. DeepSeek reversed itself completely after re-weighting the 2026 leg:
A D+6.3 generic ballot and −21.7% approval in a 2026 map that protects most Democratic incumbents produces a D net gain of 3, then a likely Democratic 2028 presidential win delivers the decisive seat in Wisconsin. — DeepSeek
[Editor's note: DeepSeek's path counts the Florida and Ohio special elections as 2026 gains, but both seats face voters again in 2028. Opus flagged the same point independently. We publish model reasoning verbatim, including where our own data card disagrees with it.]
Gemini went the other way after conceding the map problem it had underweighted, and Opus barely moved, holding the position the rest of the board drifted toward. In round one the panel sat within striking distance of the market. Reading each other pulled it to a coin flip.
Three Senators Who Could Change The Answer Without An Election
Senate control can also change because someone switches sides, and the three senators whose party loyalty has been most publicly questioned all sit in the class up in 2028.
Lisa Murkowski, the Alaska Republican, has said openly that she has considered joining the Senate Democratic caucus, calling it an "interesting hypothetical" and telling Semafor there is "some openness to exploring something different than the status quo." She has also told The Hill she does not see Democrats as much better than Republicans, so this is an open question rather than a decision.
John Fetterman has spent two years denying he will leave the Democratic Party, even as talk of a switch has trailed him through his votes for Trump Cabinet nominees and his regular Fox News appearances. The loudest test yet came on September 9, when he appeared by recorded video at the Republican midterm convention in Dallas to introduce Pennsylvania's Republican senator, Dave McCormick, as "a great American" and said in the video, "I'm always going to reject the extremes and socialism and that anti-American way of life" (ABC News). He answered the switch question the same day, in a statement: "I'm a lifelong Democrat who votes overwhelmingly with my party — and should the Senate flip in November, I will still be the 51st vote that I ran on" (AP via KSAT). He has told Spotlight PA he would formally switch only if the party made opposition to Israel part of its official platform, and the Pennsylvania Working Families Party has said it will back a primary challenger against him in 2028.
Mark Kelly's case is the sharpest of the three. In November 2025 the Pentagon opened a review of Kelly, a retired Navy captain, over a video in which he and five other Democratic lawmakers reminded service members of their duty to refuse illegal orders, and said it could recall him to active duty (CNN, NPR). Kelly called it an attempt to chill military dissent and sued Defense Secretary Pete Hegseth. In February 2026 a federal judge blocked the action, writing that the department had "trampled on Senator Kelly's First Amendment freedoms"; the government's appeal was argued before the D.C. Circuit in May 2026, and no ruling had been reported as of this update.
Every model was given the switch record as it stood on August 22, and most assigned it a small tail in both directions. None of them saw the Dallas video. Fable was blunt about the size of the effect: a formal switch by either Murkowski or Fetterman would move this market five to eight points on its own, about the size of the entire gap between the panel and the price. The Dallas video is the closest anyone has come to testing that, and the price history's answer so far is silence: not one contract has traded on this board since it aired, and the quotes that disappeared in the day after it, which nobody has tied to the video, were pulled, not hit, and were back on the screen by that afternoon.
When Does This Market Actually Get Decided?
Mostly on November 3 of this year, when the midterms fill 33 of the 67 undecided seats. Every model named that result as the thing that moves its number most:
- A Democratic Net Gain Of Four Or More Seats lowers the 2028 bar. Kimi said that case moves its number toward Democrats by eight to 12 points.
- A Net Gain Of Zero Or One pushes it hard the other way. Opus said that case takes it to Republicans at 65.
Traders currently make Republicans slight favorites to hold the Senate this year while giving Democrats the House comfortably. After the midterms, the first real 2028 presidential polling matters most, because Nevada, Pennsylvania, Arizona and Georgia all track the top of the ballot. The new Senate is seated January 3, 2029, and the market settles February 1, 2029 on the party of the Senate's President pro tempore, which the majority elects, though Kalshi can settle it early once the media calls reach a consensus. This page is re-scored when the story moves.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $188M traded
- 2027 Pro Football Champion · $69M traded
- 2028 U.S. Presidential Election winner? · $61M traded
- 2028 Republican presidential nominee · $59M traded
- Pro Baseball Champion · $58M traded
- Bitcoin price at the end of 2026 · $32M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.
The Bottom Line
This board is thin. The two 2028 Senate contracts have traded about $3,900 combined since they opened in May 2026, against roughly $8.1 million on the 2026 Senate market next door. A 57-cent print on a book that size is a handful of orders parked near the far more liquid presidential market, and those orders vanished in late August, came back a day later, vanished again this week, and came back again within the day.
Six points is not a scandal. But most of this question gets answered this November, and a market that has traded $3,900 has not really answered it at all.
Contracts on this board trade in cents and settle at either $1 or nothing, and a book this thin means the price you see may not be the price you get.
More on this: 2028 House Control Odds · 2028 Presidential Election Odds · 2026 Senate Races: Every Kalshi Board We Track · Senate Control 2026 Odds: The Four Seats That Decide It · Kalshi 2026 Senate Odds: Which Party Wins The Senate?
