Show HN: FoundationForecast: unified API for time series foundation models

Category: ai-ml

Tags: time-series, forecasting, foundation-models, unified-api, benchmarking

Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 8, Utility: 7)

FoundationForecast provides a unified Python API for time series foundation models, wrapping over a dozen models (Chronos, TimesFM, Moirai, Toto, etc.) with consistent forecast, cross-validation, and anomaly detection methods. Its key value is eliminating API fragmentation and enabling fair benchmarking, with notable extras like native exogenous variable support and CI-verified leaderboard replication for GIFT-Eval and fev-bench.

Target audience: data engineers, ML researchers, backend devs working with forecasting

Repository: https://github.com/TimeCopilot/foundationforecast · Python · Apache-2.0 · 12 stars

View on Hacker News