TL;DR
Kalshi prices a 26-cent chance the Fed HIKES next week; the eight-model AI panel says 12%. The market says zero 2026 cuts at 83 cents; the panel's modal call is two cuts. One of those gaps closes Wednesday at 2pm ET, and the other settles all year. Every verdict below was made without seeing a single market price.
The FOMC meets Tuesday and Wednesday, and the Kalshi crowd has priced an unusual meeting: a hold at 75 cents, a quarter-point HIKE at 26 cents, and effectively zero chance of a cut. With oil up roughly 14% this month on Middle East tension, the crowd is treating an inflation re-scare as live. We put the same question to eight AI models, price-blind, across every lab: what does the Fed actually do next week, and how many cuts happen in 2026 at all?
The Market
| Venue | Kalshi — a CFTC-regulated exchange for event contracts (18+; availability varies by state, as of July 2026) |
| Market 1 | "Fed decision in Jul 2026?" — five outcomes, settles on the July 29 FOMC statement |
| Market 2 | "Number of rate cuts in 2026?" — the full-year count, settles December 31 |
| Prices | Live Kalshi prices as of July 24, 2026 |
| Protocol | 8 AI models, each scoring the FULL outcome distribution in one pass, no market prices shown |
The July Decision: Hold 66%, And The Hike Is The Argument
| Outcome | Kalshi | AI blend | ChatGPT (GPT-5.5) | Claude Fable | Claude Opus | Claude Sonnet | Gemini 3.1 Pro | GLM 5.2 | Kimi K3 | DeepSeek V4 |
|---|---|---|---|---|---|---|---|---|---|---|
| Cut more than 25bps | 1¢ | 2% | 1% | 2% | 2% | 2% | 1% | 5% | 3% | 0% |
| Cut 25bps | 1¢ | 17% | 4% | 28% | 22% | 28% | 5% | 32% | 15% | 5% |
| Hold | 75¢ | 66% | 63% | 64% | 74% | 61% | 65% | 58% | 67% | 80% |
| Hike 25bps | 26¢ | 12% | 30% | 5% | 2% | 8% | 25% | 3% | 12% | 14% |
| Hike more than 25bps | 1¢ | 2% | 2% | 1% | 0% | 1% | 4% | 2% | 3% | 1% |
The headline gap: the crowd pays 26 cents for a July hike, the panel averages 12%. Claude Fable's read: "Oil's 14% spike and geopolitical uncertainty argue for patience; base rates favor hold, with a meaningful cut chance if disinflation and labor softening had momentum before the shock." The counterpoint inside the panel is real, though. ChatGPT (GPT-5.5): "Hold is most likely, but elevated inflation, oil shock risk, and firm labor conditions make a 25bp hike a live minority outcome."
There is a second, quieter gap on the other side: the panel gives a July CUT 17% on average while the market pays a single cent for it. That one is worth calling out honestly, because the crowd holds an information edge here (see The Traps).
The 2026 Cut Count: The Panel Says Two, The Crowd Says Zero
| 2026 cuts | Kalshi | AI blend | ChatGPT (GPT-5.5) | Claude Fable | Claude Opus | Claude Sonnet | Gemini 3.1 Pro | GLM 5.2 | Kimi K3 | DeepSeek V4 |
|---|---|---|---|---|---|---|---|---|---|---|
| Exactly 0 | 83¢ | 13% | 60% | 5% | 3% | 10% | 10% | 2% | 7% | 5% |
| Exactly 1 | — | 17% | 25% | 15% | 9% | 22% | 15% | 15% | 17% | 15% |
| Exactly 2 | 4¢ | 27% | 10% | 28% | 24% | 30% | 25% | 35% | 30% | 35% |
| Exactly 3 | 1¢ | 23% | 4% | 27% | 29% | 22% | 25% | 25% | 25% | 30% |
| Exactly 4 | 1¢ | 12% | 1% | 15% | 20% | 10% | 15% | 13% | 13% | 10% |
| Exactly 5 | — | 5% | 0% | 6% | 9% | 4% | 7% | 5% | 5% | 3% |
| Exactly 6 | — | 2% | 0% | 3% | 4% | 2% | 2% | 3% | 2% | 1% |
| 7 or more | — | 1% | 0% | 1% | 2% | 0% | 1% | 2% | 1% | 1% |
The market is emphatic: 83 cents on zero cuts all year. The panel's distribution centers on two (27%) with zero cuts at just 13% on average. The widest single split on the whole board sits right here, and the range on "zero cuts" inside the panel runs from 2% to 60%, which is itself the tell: the models that weighted this year's inflation re-acceleration hardest landed nearest the crowd.
The Traps
Two honesty notes before anyone acts on a gap. First, the models reason from training knowledge plus the context we hand them; the crowd trades on this week's data, speeches, and positioning. On fast-moving macro questions that information edge is real, and the cut-count market is exactly where it bites hardest. Second, the decision market settles in five days but the cut-count market runs to December 31, so the "zero cuts at 83" price also carries five months of path risk that a one-meeting market does not. When the panel and the crowd disagree this hard, someone is wrong, and our scoreboard grades every one of these verdicts when the markets settle.
The Full Reasoning
ChatGPT (GPT-5.5): Key drivers, in order: first, inflation is still above target, with recent energy pressure worsened by oil rising about 14% this month, so cuts look poorly timed. Second, the labor market appears stable rather than recessionary: unemployment is low-to-moderate, job gains have slowed but not broken, and wage pressure is not collapsing. Third, recent Fed communication has emphasized price stability and patience, with some hawkish officials likely open to tightening if inflation expectations or energy pass-through worsen. Base rates matter: the Fed usually avoids surprising at non-SEP meetings and strongly prefers 25bp increments, so hold remains the modal outcome and moves larger than 25bp are very unlikely. My main uncertainty is whether policymakers see the oil shock as temporary supply noise or a threat to inflation expectations. The single news item that would most change my estimate is a fresh inflation-expectations or CPI/PCE-related release showing broad, non-energy reacceleration before the meeting.
Claude Fable: Key drivers, in order: (1) the oil shock , a 14% monthly crude move feeds headline CPI within weeks and historically makes the Fed pause rather than ease into an inflation impulse, even one it views as transitory supply-side; (2) recent Fed communication norms , the FOMC telegraphs moves, and absent a clearly signaled July cut in June's dots/pressers, an unsignaled move is rare (holds occur at roughly 70-80% of meetings across the post-2000 era); (3) labor-market trajectory , if unemployment has been drifting up, the employment mandate keeps a 25bp cut live as the clear second scenario. Hikes are tail outcomes: the bar to tighten on a supply-driven energy spike is high post-2022 lessons. I'm most uncertain about how much easing the Fed had already signaled entering July , that alone swings hold vs. cut_25bps by 20+ points. The single most decisive news would be the June core CPI/PCE print (or a Powell/vice-chair speech in the blackout run-up) confirming or breaking the disinflation trend.
Claude Opus: Weights, in order: (1) Base rate , the FOMC holds at the large majority of meetings, and holds are especially likely when the committee has not clearly pre-signaled a move in its communications. (2) The oil shock: a ~14% monthly crude move on Middle East tension pushes headline inflation up and raises the risk of expectations drift, which historically makes the Fed wait for confirmation that it is a relative-price shock rather than cut into it. (3) Labor market: cuts would require visible softening , rising unemployment or negative payroll revisions , and a cut is the far more plausible non-hold direction, so I put ~22% on 25bp versus ~2% combined on hikes. Hikes are near-tail: the Fed does not reverse direction on a supply shock absent an expectations breakout. My largest uncertainty is the labor data I cannot see , the state of the last two payroll and claims prints, and whether officials have already telegraphed a July move. The single most decision-changing news would be a sharply weak July payroll/unemployment print (or a Powell-level signal that the oil move is being looked through), which would push the cut probability well above 50%.
Claude Sonnet: Weighting: (1) a multi-quarter easing cycle already in motion creates default momentum toward a cut absent a shock; (2) the ~14% oil spike from Middle East tension is a fresh, acute inflationary supply shock that historically makes committees pause rather than cut, since headline CPI/PCE will tick up mechanically over the next 1-2 prints; (3) labor market trajectory is the swing factor , if payrolls/claims have been softening, that argues for cutting through the noise, but oil shocks specifically raise fear of second-round effects on inflation expectations, which the Fed treats as more dangerous than growth risk. Base rates: FOMC hikes are rare outside acute inflation surprises, so tails stay small. Most uncertain about: how much labor-market deterioration has occurred since the last meeting , if it's sharp, cut_25bps could be underweighted. The single most-moving datapoint: the next CPI/PCE print's core (ex-energy) reading, which reveals whether the oil shock is passing through to core prices.
Related Verdicts
Model estimates generated July 24, 2026, price-blind. 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. Kalshi is a CFTC-regulated exchange; 18+, availability varies by state.
FAQ
When does the July decision market settle?
The FOMC statement lands Wednesday, July 29, 2026 at 2pm ET; the decision market settles on the announced action.
Why do the models disagree with the market so much on cuts?
The market trades on live 2026 data; the models reason from training knowledge plus stated context. When they disagree, settlement tells us whose process wins, and we grade every verdict on our scoreboard.
Are model verdicts betting advice?
No. Model verdicts are model estimates, not betting or financial advice. Treat them as one input among many and make your own decisions.
