Training camps opened this week, which means the 2026 NFL season is officially a market. Kalshi lists all 32 teams to win Super Bowl LXI, and we put the question to seven AI models — price-blind, so none of them could lean on the crowd. Below is the full board: every team, every model, the blended verdict, and the gap between the machines and the money. (New to Kalshi's cent pricing? Our walkthrough of NFL betting on Kalshi covers how the contracts work.)
The Market
| Venue | Kalshi — a CFTC-regulated exchange (18+; availability varies by state, as of July 2026) |
| The event | "2027 Pro Football Champion" — 32 yes/no contracts, one per team |
| Prices | Live Kalshi last-trade prices as of July 24, 2026 |
| Protocol | 7 AI models, each scoring every team WITHOUT seeing any market price |
Free: 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. One email, Sundays. One email, Sundays — the signup box is at the bottom of this page, or just keep reading: the record it promises is on this one.
How To Read This Board
Each model estimated every team's championship probability independently, seeing rosters, coaching changes, and offseason moves, but no market prices. A positive gap means the AI panel is more bullish than the Kalshi crowd; negative means the crowd likes the team more than the machines do. In a 32-team field, single digits are normal and double digits are conviction. The two conference paths get boards of their own as well: the AFC championship market and the NFC championship market.
Where The Machines Split From The Money
L.A. Rams: market 15¢, AI blend 5% (10 points below the crowd). DeepSeek V4: "Rams have strong coaching with McVay but an aging Stafford and post-Donald defense; uncertain QB situation by 2026 limits their odds."
Seattle: market 8¢, AI blend 4% (4 points below the crowd). GLM 5.2: "Seahawks have a solid roster but quarterback uncertainty with Sam Darnold and a competitive NFC West make them a mid-tier contender with longshot championship odds."
Philadelphia: market 5¢, AI blend 9% (4 points above the crowd). Claude Opus: "Philadelphia remains a loaded, well-coached Super Bowl contender, but a ~30-team field caps even elite favorites."
Kansas City: market 7¢, AI blend 10% (4 points above the crowd). Claude Opus: "Kansas City remains an elite perennial contender with Mahomes and Reid, but repeat variance and a deep field cap the odds."
The Full Board: 32 Teams, Seven Models, One Blend
| Team | Kalshi | AI blend | Gap | ChatGPT (GPT-5.5) | Claude Fable | Claude Opus | Claude Sonnet | GLM 5.2 | Kimi K3 | DeepSeek V4 |
|---|---|---|---|---|---|---|---|---|---|---|
| Kansas City | 7¢ | 10% | +4 | 8% | 8% | 15% | 8% | 11% | 10% | 13% |
| Philadelphia | 5¢ | 9% | +4 | 8% | 11% | 13% | 6% | 8% | 8% | 7% |
| Buffalo | 8¢ | 8% | +0 | 8% | 9% | 11% | 4% | 8% | 10% | 8% |
| Baltimore | 7¢ | 8% | +1 | 8% | 9% | 11% | 4% | 7% | 9% | 8% |
| Detroit | 4¢ | 6% | +2 | 7% | 7% | 8% | 4% | 6% | 4% | 9% |
| San Francisco | 5¢ | 6% | +1 | 6% | 8% | 8% | 5% | 6% | 5% | 7% |
| Green Bay | 4¢ | 5% | +1 | 4% | 5% | 7% | 5% | 5% | 6% | 4% |
| L.A. Rams | 15¢ | 5% | -10 | 4% | 7% | 8% | 6% | 4% | 5% | 3% |
| Washington | 3¢ | 5% | +2 | 4% | 2% | 5% | 5% | 8% | 5% | 4% |
| Houston | 4¢ | 5% | +1 | 6% | 4% | 5% | 5% | 4% | 5% | 6% |
| Cincinnati | 5¢ | 5% | -0 | 4% | 4% | 4% | 3% | 4% | 5% | 9% |
| Denver | 5¢ | 4% | -1 | 4% | 7% | 7% | 3% | 3% | 4% | 2% |
| L.A. Chargers | 5¢ | 4% | -1 | 4% | 4% | 4% | 2% | 4% | 4% | 4% |
| Seattle | 8¢ | 4% | -4 | 2% | 4% | 2% | 4% | 2% | 9% | 2% |
| Dallas | 4¢ | 4% | -0 | 4% | 2% | 5% | 4% | 5% | 3% | 4% |
| New England | 4¢ | 4% | -0 | 1.8% | 4% | 3% | 3% | 3% | 8% | 2% |
| Minnesota | 2¢ | 3% | +1 | 4% | 2% | 4% | 3% | 2% | 3% | 2% |
| Chicago | 4¢ | 3% | -1 | 2% | 1.5% | 3% | 1.2% | 4% | 3% | 2% |
| Jacksonville | 3¢ | 2% | -0 | 2% | 3% | 2% | 3% | 1.5% | 4% | 2% |
| Tampa Bay | 2¢ | 2% | +0 | 2% | 2% | 3% | 3% | 3% | 2% | 1.7% |
| Atlanta | 1¢ | 2% | +1 | 1.8% | 2% | 2% | 3% | 2% | 2% | 2% |
| Miami | 1¢ | 2% | +1 | 3% | 1.5% | 2% | 3% | 2% | 2% | 2% |
| Pittsburgh | 2¢ | 2% | +0 | 2% | 1.5% | 3% | 2% | 2% | 2% | 2% |
| Indianapolis | 2¢ | 2% | +0 | 1.8% | 2% | 2% | 2% | 2% | 2% | 4% |
| N.Y. Jets | 1¢ | 2% | +1 | 1.8% | 1.2% | 1.8% | 4% | 4% | 1.0% | 1.5% |
| N.Y. Giants | 2¢ | 1.7% | -0 | 1.2% | 1.2% | 1.2% | 3% | 1.5% | 1.5% | 2% |
| Arizona | 1¢ | 1.6% | +1 | — | 1.5% | 1.2% | 2% | 1.5% | 2% | 1.5% |
| Las Vegas | 1¢ | 1.5% | +0 | 1.2% | 1.0% | 1.2% | 3% | 1.5% | 1.5% | 1.0% |
| Cleveland | 1¢ | 1.3% | +0 | 1.2% | 1.0% | 1.2% | 3% | 1.5% | 0.5% | 1.0% |
| New Orleans | 1¢ | 1.3% | +0 | — | 0.8% | 1.0% | 2% | 1.5% | 0.8% | 1.5% |
| Carolina | 1¢ | 1.1% | +0 | — | 0.8% | 1.2% | 2% | 0.8% | 1.2% | 0.7% |
| Tennessee | 1¢ | 0.7% | -0 | — | 0.5% | 0.8% | 0.5% | 0.8% | 0.8% | 0.8% |
Every seat on this panel is graded against real market settlements — records to date: GPT 83% on 1,693 graded calls · Claude Fable 85% on 331 graded calls · Claude Opus 86% on 382 graded calls · Claude Sonnet 84% on 363 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.
Model estimates generated July 24, 2026. These are model estimates, not predictions of fact and not financial or trading advice. Models are frequently wrong; the market price reflects real traders' money.
The Regular-Season Version
The regular-season version of this question is its own board: NFL win totals 2026, all 32 ladders priced against the panel's expected-wins model.
FAQ
How can the model percentages differ so much from the Kalshi price?
The models never see the price — that's deliberate. When they disagree with the crowd, one side is wrong, and tracking which side wins over time is exactly what our model scoreboard measures.
Why don't the percentages add up to 100%?
Each model scores every team independently, so small over- or under-shoots accumulate across 32 teams. The relative order and the gaps matter more than the raw sum.
Are model verdicts betting advice?
No. Model verdicts are model estimates of how a market might resolve, not betting or financial advice. Treat them as one input among many and make your own decisions.
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.
