For temperature-market analysts

The station is the edge.

Daily temperature markets don't settle on "the weather in Paris." They settle on one named airport gauge — and Paris's gauge runs 1–3°C cooler than the city your weather app shows. Voltcast gives you probability distributions on the exact station, in the exact 1°C buckets markets trade — verified against METAR observations every single day.

Analytics only — never wagering advice. Free tier available; no card required to look around.

Edge 1 · Resolution

The right gauge, not the right city

Paris markets settle on Le Bourget — swapped from Charles-de-Gaulle in April 2026 after a sensor affair that ended in a Météo-France criminal complaint. London alternates between two airports per series. Trading the city-center number is the most common structural mistake in these markets. Our resolution-station registry names every gauge and flags the ones that stop reporting.

Edge 2 · Distribution

Probabilities in market buckets

A point forecast ("34° tomorrow") is useless when the market pays by the bucket. We publish the full distribution — P05 to P95 and per-1°C-bucket probabilities — built from the 51-member ECMWF ensemble, bias-corrected per station against 35 years of history, for every settlement gauge, every day, up to 5 days out.

Edge 3 · Uncertainty

Know when NOT to trust anyone

Where the ensemble's 51 members agree, tomorrow is priced in. Where they disagree is where forecasts — everyone's — get wobbly, and where bucket prices move hardest. Our uncertainty layer shows the disagreement per station, live.

Tomorrow, right now

all 43 stations →

Live from the same feed subscribers query — the market cities, tomorrow's Tmax distribution, top buckets.

Loading live distributions…

We publish our scores. Even the bad ones.

Every forecast we sell is scored daily against real METAR observations, in public, on /accuracy — CRPS and bucket Brier, next to the raw-ensemble baseline it has to beat to even serve. When we lose, you see it (we write it up). No other data vendor in this niche shows you their misses. That's the point: you're pricing risk — you need a data source whose error bars you can trust.

Which one is you?

For market analysts

Volt Temp €99/mo

  • Station-exact Tmax/Tmin distributions, D+1 to D+5
  • Per-1°C bucket probabilities (the market's shape)
  • Resolution-station registry + integrity status
  • CME-city degree-day settle projections
  • Population-weighted zone temperature
  • Webhooks · JSON API · Python/TS SDKs
Start with Volt Temp

Quant €999/mo

For systematic traders

  • The Voltcast Desk — live station trajectories, run drift, event tape
  • Everything in Volt Temp + the full power-market API
  • Point-in-time-correct training matrices per station — the exact rows our own model eats
  • As-of reconstruction (what was knowable when)
  • Full-history parquet exports · 3000 req/min
Build on Quant

Balancing €299/mo

For energy desks

  • Weather is also the #1 power-price driver — same data, energy angle
  • Imbalance prices (7 markets, revision-aware) + forecasts
  • FCR/aFRR/mFRR reserve prices · outages
  • Wind/solar forecasts + zone temperature
Start with Balancing

All self-serve via Stripe · annual −15% · cancel anytime · free tier to explore first

Voltcast provides data and analytics for informational purposes only — never financial, trading or wagering advice, and never a recommendation to take any market position. Prediction-market participation is restricted in many jurisdictions; verify your own eligibility and the named resolution source in each market's rules. Forecast inputs: Open-Meteo / ECMWF (CC BY 4.0); observations: NOAA METAR; climatology: NOAA ISD.