TL;DR
The September Fed board is the most aggressive macro price on Kalshi: a quarter-point HIKE is the 50-cent favorite, with a hold at 44. The panel — given the fetched CPI and payrolls history but told Wednesday's July decision was still unknown — blends the September hike at 7%. The crowd is trading an oil-shock inflation regime; the models keep citing June's negative core print and a labor market that added just 57K jobs. Wednesday's decision and statement will move this board immediately, and we will grade every number here against the September 16 settlement.
The Distribution: September 2026 FOMC
| 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¢ | 14% | 4% | 6% | 5% | 13% | 25% | 28% | 10% | 20% |
| Cut 25bps | 4¢ | 48% | 24% | 55% | 55% | 45% | 60% | 40% | 46% | 55% |
| Hold | 44¢ | 30% | 48% | 36% | 37% | 32% | 10% | 25% | 35% | 20% |
| Hike 25bps | 50¢ | 7% | 22% | 2% | 2% | 8% | 5% | 4% | 7% | 4% |
| Hike more than 25bps | 3¢ | 1% | 2% | 0% | 0% | 2% | 0% | 3% | 2% | 1% |
Generated with the July 29 decision still unknown; every model conditioned on the plausible July paths.
The Reasoning
ChatGPT (GPT-5.5): I condition July mostly on a hold, with a meaningful chance of a 25 bp hike and only a small chance of a cut. If July holds, September is likely a data-dependent choice: the six-month core CPI run rate has cooled, especially with June slightly negative, and payroll growth has decelerated sharply, so a 25 bp cut becomes plausible if July/August labor or inflation data soften further. But a July oil shock is large enough to keep the Committee cautious about inflation expectations and headline pass-through, making hold more likely than cut. If July hikes, September most likely becomes a pause unless inflation keeps worsening; a second hike is possible but less likely given labor fragility. If July cuts, it would probably reflect acute concern about growth, raising follow-on cut risk. Moves larger than 25 bp are low-probability without financial stress or a sudden inflation shock.
Claude Fable: Disinflation is clear: core CPI decelerated from ~0.3% MoM early in the year to +0.21% in May and -0.02% in June, an outright negative print. Labor is softening too , payrolls slid from +214K in March to +57K in June, with a negative February. This combination pushes the Fed toward easing. The July 29 meeting is the wildcard: if the Fed cuts in July (plausible given June data), September could be a hold-to-assess or a follow-on cut; if July is a hold with dovish guidance, September becomes the likely first cut. Either branch leaves substantial probability on a 25bp September cut. The ~30% July oil rally is the main hawkish risk, but the Fed typically looks through energy shocks for core; it mainly supports the hold scenario rather than hikes. A 50bp cut requires visible labor-market deterioration in the July/August reports , possible but not baseline. Hikes are nearly ruled out by the data trajectory.
Claude Opus: The disinflation signal is real: core CPI has stepped down from +0.30%/+0.38% early in the year to +0.21% in May and -0.02% in June, pulling the 3-month annualized run rate near or below target. Labor is the stronger cut argument , payrolls have decayed from +214K in March to +129K and then +57K, with February's -156K showing the trend is fragile rather than merely normalizing. That combination is the textbook setup for easing. The main complication is timing and oil: a ~30% July crude rally will push headline inflation up and may bleed into core via transport and airfares with a lag, giving hawks a reason to wait for confirmation. If the Fed cuts July 29, September becomes more likely a hold (back-to-back cuts are less common absent crisis), which is why hold retains substantial weight. A 50bp cut requires a labor shock not yet in the data, and hikes are nearly excluded with core inflation flat and hiring stalling.
Model estimates generated July 27, 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.
Related Verdicts
FAQ
Why do the model percentages differ from the Kalshi price?
The models never see the price. When they disagree with the crowd, one side is wrong, and we grade every verdict against real settlements 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.
