Kalshi temperature markets, priced properly.
Calibrated probabilities and a live, fee-adjusted edge for Kalshi’s daily highest-temperature brackets — aware of the exact NWS station, the LST settlement window, and the fee drag that kills cheap-tail bets.
Free during open beta · NWS-station aware · net-of-fee EV.
New York
Highest temp · today- ≤ 84° 6%3%
- 85–86° 14%9%
- 87–88° 31%46% + 15
- 89–90° 33%31%
- 91–92° 12%9%
- ≥ 93° 4%2%
87–88° is underpriced. Model fair value is 46% against a 31% market. Invalidate if the high prints before 2pm local.
Illustrative. Live in the terminal.
NWS-station & settlement aware
Kalshi settles on the next-day NWS Climate Report in Local Standard Time — a different source and clock than Polymarket. Temprr knows the gauge, the LST window, and the DST shift that breaks naive midnight math.
Fee-adjusted EV per bracket
A $0.01 fee on a $0.05 contract is a 20% tax. Temprr surfaces net edge after spread and fees per bracket, and flags the cheap-tail death zone before you buy in.
Calibrated, not a single model
A multi-model, fat-tail-aware distribution over Kalshi’s temperature brackets — the layer that turns a forecast into a probability you can actually price against.
The same day can settle 1°F apart on Kalshi vs Polymarket.
Temprr shows both settlement sources side by side, with a boundary-divergence flag — so you can hedge or arbitrage the gap instead of getting caught by it.
Compare the two →For the Kalshi trader
Everything the Kalshi temperature trade needs.
Calibrated brackets, net-of-fee edge, station awareness and a wallet read — without juggling a dozen tabs.
Right station, every time
Every market is anchored to its exact settlement gauge — NYC on KLGA, Paris on Le Bourget, not the city center. The #1 silent edge-killer, designed out by default.
Calibrated, not Gaussian
A fat-tail-aware bucket distribution with per-station bias correction and a visible uncertainty band — not a naive bell curve that blows up on the tails.
The edge, surfaced
Model fair value vs live market price, rendered as an explicit per-bucket divergence. The one comparison no competitor puts on screen.
Forecast it, not race it
Calibrated fair value recomputed every 60 seconds against live observations, with freshness in concrete minutes. A view of where the day will finish while the high is still open.
Net edge after fees
We surface whether a divergence actually clears the spread and fee hurdle before you trade — steering you away from the fee-toxic cheap tail.
It shows its work
Evidence chain, settlement-station label, calibration sample size and a live freshness dot. Methodology as the trust signal — no black box, no hype.
Stop guessing the daily high. Know it.
Temprr is in open beta and free. Create an account and start today.
Free during open beta · members-only.