The weather is priced wrong.
Temprr shows you where.
A calibrated weather-trading terminal for Polymarket and Kalshi daily-temperature markets. Official station observations and multi-model forecasts, turned into a live edge against the market price.
- ✓ Free at launch
- ✓ ~51 cities
- ✓ No API key
- ✓ Paste a public wallet for P&L
- ≤ 84° 6% 3%
- 85–86° 14% 9%
- 87–88° 31% 46% ▲ + 15
- 89–90° 33% 31%
- 91–92° 12% 9%
- ≥ 93° 4% 2%
Illustrative. Live in the terminal.
Why traders lose edge
The market is priced wrong. The hard part is seeing where.
The edge is real — on-chain leaderboards prove it. But it leaks out through five recurring mistakes. Temprr is built to close each one.
“I was providing exit liquidity for the bots that got there first.”
Amateurs poll forecasts every 15–60 minutes. The repricing window is now 5–15.
Temprr recomputes fair value every 60s against live obs and overlays the live book — so you see divergence the moment it opens, not after.
Trading the wrong station silently loses trades.
KLGA runs 3–5°F cooler than midtown; the gauge even differs between series.
Every market is pinned to its verified settlement station. Your obs, forecast and buckets all reference the gauge that actually settles.
“Weather doesn’t follow a bell curve. My model did.”
A “2-sigma” event that should hit 5% of the time actually hits 10–12%.
DEB blends multiple models into a calibrated, fat-tail-aware distribution, validated against history — the empirical-error layer DIY models skip.
The picture is scattered across a dozen tabs and timezone math.
NWS, Weather.com, Ventusky, Windy, METAR — plus Zulu-to-local conversions.
One terminal: official obs + multi-model forecast + calibrated bucket distribution + live odds + peak-window, in the station’s local time.
The terminal
Everything the trade needs, on one screen.
Temprr replaces a dozen browser tabs and a spreadsheet with a single, calibrated, evidence-first view of every daily-temperature market.
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.
See it first, not race it
Calibrated fair value recomputed every 60 seconds against live observations, with freshness in concrete minutes. Spot divergence the moment it opens.
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.
How it works
From forecast to a priced, tradable edge.
Anchor to the gauge that settles
Pick a city. Temprr pins the market to its exact settlement station and pulls live official observations — METAR/TAF plus native networks (JMA, HKO, CWA, KNMI, Météo-France).
Calibrate the distribution
DEB blends multiple forecast models into a fat-tail-aware probability over the market’s exact temperature buckets, bias-corrected per station and validated against history.
Surface the edge — and the rule
Live Polymarket odds overlay the model. Temprr shows the divergence, whether it clears the spread, and the invalidation rule that kills the trade.
Global coverage
Built for Polymarket and the world — not four US cities.
~51 cities across the Americas, Europe, the Middle East and Asia, each wired to its native official feed — JMA, HKO, CWA, KNMI, Météo-France, SMN — where durable edge still lives and the bots are sparse.
Americas
16Europe & Middle East
13Asia-Pacific
14Who it’s for
For everyone who knows the weather is tradable.
The model-checker
You pull GFS/ECMWF and METAR by hand and keep a spreadsheet. Temprr is that spreadsheet, calibrated, live, and right about the station.
The quant-curious builder
You cloned a GFS-ensemble bot or read the “$1k → $24k” thread. Skip the data plumbing — start from a correct, calibrated layer.
The cross-platform arbitrageur
You trade Kalshi and Polymarket. See both settlement sources side by side and the 1°F boundary cases that flip a contract.
The curious crossover
You came via Robinhood-Kalshi or a Polymarket meme. Plain-language reads, the right station, and guardrails before you size up.
FAQ
Questions, answered.
What is Temprr? +
Temprr is a weather-intelligence trading terminal for daily-temperature prediction markets on Polymarket and Kalshi. It turns official station observations and multi-model forecasts into a calibrated probability over each market’s temperature buckets, then overlays the live market price to show where the two diverge.
Which markets does it cover? +
Daily highest/lowest temperature markets across roughly 51 cities worldwide — the Americas, Europe, the Middle East and Asia-Pacific — with native official feeds per region (JMA, HKO, CWA, KNMI, Météo-France, SMN) on top of METAR/ASOS.
Is it free? +
Temprr launches free and members-only. No payment, no API key required. Join the waitlist for an early seat.
Does Temprr place trades for me? +
No. Temprr is an analytics and decision terminal — it surfaces the edge, the evidence and the invalidation rule. You trade on Polymarket or Kalshi yourself.
Get in before the book does.
Temprr is in private beta. Join the waitlist for an early, free members-only seat.