Benchmark-verified demand forecasts for U.S. power markets.
Gramm forecasts how much electricity the grid will need, sharper than each operator’s own forecast. 1.56–2.82% error across 7 U.S. markets.*
Illustrative savings by utility footprint.
- Major ERCOT LSE$4.2M/yr· ERCOT
- Texas retailer$3.8M/yr· ERCOT
- ERCOT utility$3.8M/yr· ERCOT
- Mid-Atlantic IOU$2.1M/yr· PJM
- Chicago-area IOU$3.4M/yr· PJM
- Midwest IOU$2.6M/yr· MISO
- Gulf-area LSE$2.4M/yr· MISO
- Michigan utility$2.1M/yr· MISO
- Northern CA LSE$1.8M/yr· CAISO
- Virginia IOU$3.1M/yr· PJM
- Southern CA LSE$1.6M/yr· CAISO
- Upstate NY IOU$1.8M/yr· NYISO
- NYC-area IOU$2.1M/yr· NYISO
- New England IOU$1.2M/yr· ISO-NE
- SPP utility$1.2M/yr· SPP
- Plains IOU$1.1M/yr· SPP
Illustrative only — MAPE reduction × avg hourly load × typical imbalance penalty × 8,760 hours. Past performance is not a guarantee of future results. Methodology · Disclaimer
Public benchmark
Territory-level accuracy, with the baseline visible.
Day-ahead MAPE over a 12-month holdout window. Each region measured against its own published ISO baseline.
Live API
One request, full forecast.
Bearer token auth, JSON response, prediction intervals included. Day-ahead ERCOT forecast in 47ms.
View API docsNext step
Backtest, or start building.
Run a side-by-side backtest
Get your market and horizon benchmarked against the feed you use today. One business day turnaround.
Request backtestGo live
Free API key, no credit card. Bearer token auth, JSON response. Most teams are in production the same day.
View quickstartTry the API.
100 requests/hour and 1,000 requests/day on the free tier. No credit card.
