Jeanne Shaheen's retirement opened the New Hampshire Senate seat she won in 2008, and Kalshi's market on it has made up its mind: the Democratic side trades at 88 cents, the Republican side at 12. Then we ran the race through our eight-model, price-blind panel, armed with every public poll of the matchup and the midterm history ledger, and got the largest market-versus-model break of this batch: the panel blends the Democrats at 67.9%. Same favorite, very different confidence, and the argument between those two numbers is the whole race.
The Quick Answer
Panel and market agree on the direction and split hard on the size: Democrats 67.9% against an 88-cent price, Republicans 32.1% against 12 cents. Every seat named the same tension: the polls show likely nominee Chris Pappas leading John E. Sununu by only 3 to 6 points and tightening since April, which reads more like Lean D than the near-lock the market is charging for. This is an open seat, and the Republican recruit has won this exact seat before. The full reasoning, the poll-by-poll file, and the one poll that fools everyone are below.
New to event markets? Start with how election markets work.
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
Prices on this page refresh as the race develops; every figure carries its fetch date, and the board re-scores after the September 8 primaries. Prices as of August 7, 2026.
The Board
| Outcome | Market Price | AI Blend | Gap (blend% − mkt¢) |
|---|---|---|---|
| Democratic party | 88¢ | 67.9% | −20.1 |
| Republican party | 12¢ | 32.1% | +20.1 |
The two contracts pay on the party sworn in, not on any named candidate, so an independent winner would resolve both No; the panel's numbers sum to 100 because no seat priced that tail above rounding.
The case for each side of the New Hampshire (NH) Senate race, priced on Kalshi: Why Chris Pappas (D) Will Win the NH Senate Primary · Why Karishma Manzur (D) Will Win the NH Senate Primary · Why John E. Sununu (R) Will Win the NH Senate Primary · Why Scott Brown (R) Will Win the NH Senate Primary
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $172M traded
- 2028 U.S. Presidential Election winner? · $57M traded
- 2027 Pro Football Champion · $56M traded
- Pro Baseball Champion · $47M traded
- More tech layoffs in 2026 than in 2025? · $31M traded
- Bitcoin price at the end of 2026 · $30M 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 Market At A Glance
| Venue | Kalshi |
| The Event | "New Hampshire Senate winner?" (Kalshi event SENATENH-26), 2 party contracts, roughly $159,000 traded |
| The Contract | "If a representative of the [party] is sworn in as a Senator of New Hampshire for the term beginning in 2027, then the market resolves to Yes." Party, not person: primary upsets do not void anything. |
| The Race | Open seat; Sen. Jeanne Shaheen (D) is retiring. Primaries are September 8: Rep. Chris Pappas leads the Democratic field 53-29, former Sen. John E. Sununu leads the Republican field 59-20. Election day is November 3, 2026. |
| Settles | At the swearing-in for the term beginning January 2027 |
| Prices As Of | August 7, 2026 (candlestick closes) |
What The Models Were Given
The card carried the complete polling file from the deterministic feed, with a hard freshness audit: the newest nonpartisan public poll of the actual candidates is June 25, and the panel was told so plainly. The file: UNH in June has Pappas over Sununu 47-44, tightened from 49-42 in April; Saint Anselm in June has it 47-41, and Pappas over Scott Brown 48-36. The race registry adds the structure: one of only three Democratic-held competitive seats on the 2026 map, rated Lean D, in a state whose federal races have leaned Democratic while staying a swing state downballot. The card also carried the midterm ledger, the president's party losing ground in most postwar midterms, with the Trump-era wrinkle that midterm polling ran nearly unbiased (2018 and 2022 both within a point). And it carried a trap defused: a widely shared Quantus poll showing "Sununu" ahead 53-44 tested Chris Sununu, the former governor, who is not a declared candidate. The Republican running is his brother John E. Sununu, who held this exact seat until Shaheen took it from him in 2008.
Every Seat's Number
The blend is the equal-weight mean of the eight models' round-two numbers, generated after each model read the others' anonymized round-one reasoning and could revise. All eight seats returned in both rounds.
| Outcome | Market | Fable | Opus | Sonnet | GPT | Gemini | Kimi | GLM | DeepSeek | Blend |
|---|---|---|---|---|---|---|---|---|---|---|
| Democratic party | 88¢ | 73.0 | 78.0 | 63.0 | 64.0 | 74.0 | 64.0 | 63.0 | 64.0 | 67.9% |
| Republican party | 12¢ | 27.0 | 22.0 | 37.0 | 36.0 | 26.0 | 36.0 | 37.0 | 36.0 | 32.1% |
The panel: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT, Gemini Pro, Kimi, GLM, and DeepSeek, each pricing both sides in one pass on August 7, 2026, without seeing any prices.
Every seat on this panel is graded against real market settlements — records to date: Claude Fable 84% on 334 graded calls · Claude Opus 85% on 385 graded calls · Claude Sonnet 83% on 366 graded calls · GPT 83% on 2,013 graded calls · Gemini 85% on 2,062 graded calls · Kimi 83% on 2,099 graded calls · GLM 81% on 2,138 graded calls · DeepSeek 79% on 2,145 graded calls. Recomputed daily; the full scoreboard is public.
More live boards from the same panel: South Carolina Senate special · Darline Graham 48¢ · Maine Senate 2026 · Collins 37¢ · Michigan Senate 2026 · El-Sayed 58¢ · 2026 Senate map hub
The Democratic Case: Structure Everywhere You Look
Every seat started from the same three layers. The environment: most postwar midterms punish the president's party, and no seat found the exception conditions (a backlash cycle, a rally cycle) present. The state: New Hampshire's federal races have leaned Democratic for a decade, what Gemini Pro called an "entrenched partisan floor that protects Democratic federal candidates here," distinct from the state's swingy downballot habits. The candidate: Chris Pappas is a statewide-tested sitting congressman leading every published matchup, 47-44 and 47-41 in June, and by double digits against the Republican field's second option. Claude Opus, the panel's high seat at 78, added the polling-quality point: the Trump-era midterm polls in the card's error archive ran nearly unbiased, "so no reflexive presidential-year correction applies to a mid-single-digit lead."
The Republican Case: An Open Seat and a Man Who Has Won It
The panel's Republican number is nearly three times the market's, and the reasons are concrete. This is an open seat, not an incumbent defense; Shaheen's personal brand retires with her. John E. Sununu is, as Claude Sonnet put it, "a live threat if the environment shifts," a legacy New Hampshire name who has already won this seat once, with a consolidated primary behind him. The polls are close for a market this lopsided: a 3-point June lead is inside ordinary polling error, and the April-to-June movement ran toward the Republican. Four seats independently landed at 36-37% for exactly that bundle. What none of them found was a reason to call the race a toss-up; what all of them found was a reason to call 12 cents cheap.
The Poll That Fools Everyone
The card's most useful line may have been a correction. A Quantus survey circulating this summer shows "Sununu" leading Pappas 53-44, the only result anywhere with the Republican ahead, and it does not describe this race: it tested Chris Sununu, the former governor and the state's most popular Republican, who has not entered. Claude Fable flagged it unprompted as the thing "most people get wrong" about this race. If the market's 88 cents is partly a bet that the real Sununu polls like his brother, the panel's answer is that the actual candidate trailed in every nonpartisan actual-candidate survey in the file; the lone tie-adjacent number is a Republican committee's own internal.
Where The Panel Changed Its Mind
The revision round produced this batch's best argument arc. Gemini Pro made the largest move of any seat this week, raising its Democratic number from 66 to 74 after two peers corrected its reading of the environment: the midterm penalty "acts as a tailwind for the Democrats rather than a headwind," since the presidency it punishes is Republican. GLM moved the opposite direction inside the same debate, holding its 63 after re-examining the tightening story: the narrowing "is largely a UNH-specific movement," it wrote, since Saint Anselm's June poll still showed Pappas up 6, and "two data points from one pollster do not establish a trend." Claude Opus held its 78 as the acknowledged high outlier, noting the whole downward case "reduced to the same three inputs I priced explicitly." Nobody moved toward the market's 88.
What Would Change The Panel's Mind
- The September 8 primaries. Confirmation of Pappas-Sununu is priced; an upset on either side is not, and the first post-primary poll of the settled matchup is the single number every seat named as its trigger.
- A post-Labor Day nonpartisan poll inside 2 points. Claude Fable's stated threshold for pushing the Republican side toward 35-40.
- A generic-ballot shift. The environment argument runs through the national numbers; a sustained Republican move there erodes the structural layer of the Democratic case.
- Any new nonpartisan poll of the settled matchup. One newer partisan number already exists: a Republican-sponsored internal (Peak Insights for the NRSC, conducted June 24-27, released in July) has the race 43-42 Pappas. Internals released by a party committee earn a discount by construction, but its direction matches the tightening story, and it is the opposite of evidence for the market's 88 cents.
Settlement Timeline
| Date | What Happens |
|---|---|
| September 8, 2026 | Both primaries; the hypothetical matchups become a real one |
| September-October 2026 | Post-primary polling of the settled race; debate season |
| November 3, 2026 | Election day |
| January 2027 | Swearing-in resolves the market on the winner's party |
This page re-scores after the primaries and again as the general develops; prices and blends above are as of August 7, 2026.
How To Read These Prices
The trade the panel's numbers imply is uncomfortable precisely because the direction agrees with the market: buying the Republican side at 12 cents pays about 7.3-to-1 against a panel probability near 1-in-3, one of the widest value gaps our panel has printed on any live election board. The discipline point cuts both ways: a 20-point break between model and market on a race with a June polling file means someone is wrong, and markets have been right against our blends before, which is why every one of these calls is graded in public once it settles. The model verdict scoreboard carries that running record. Both contracts park capital until January 2027, a real cost on a 12-cent position.
More on this: 2026 Senate Races: Every Kalshi Board We Track · Why Chris Pappas (D) Will Win the NH Senate Primary · Why John E. Sununu (R) Will Win the NH Senate Primary · Why Kurt Alme (R) Will Win the Montana Senate Race · Why Ashley Moody (R) Will Win the Florida Senate Primary
The Bottom Line
The market and the panel agree New Hampshire leans Democratic and disagree about how hard. Eighty-eight cents prices a near-safe seat; the public record, three-point June polls, an open seat, a Republican who has won here before, and a tightening trend that may or may not be real, prices something looser, and the panel lands at 67.9. The primaries on September 8 start answering which read is right, and this page re-scores when they do. 18+, eligible states only.
Verdicts generated August 7, 2026, price-blind, revised once after an anonymized cross-read. Prices as of August 7, 2026 (Kalshi candlesticks). Panel: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT, Gemini Pro, Kimi, GLM, DeepSeek.
Prices on this page are Kalshi's book. If you also trade on Polymarket, code OS4 gets new users a $50 trading bonus on a $10 deposit — affiliate link; terms as stated by Polymarket; 18+, availability varies by state.
