ParaPulse

ParaPulse is an AI model analytics tool that tracks momentum across open-source models and datasets on Hugging Face, so ML practitioners can see which models are rising or fading and how fast. Instead of showing raw download counts alone, it computes growth rates and momentum-weighted rankings. It takes daily snapshots of 30,000+ models and 12,000+ datasets, then calculates 7-day, 30-day, and composite growth scores to surface trending and declining projects. Concrete features include trend charts over 7d/30d/90d windows, side-by-side multi-model comparison, momentum-weighted leaderboards, category browsing (text generation, image generation, speech, and more), a searchable feed of recently updated models and datasets, weekly digest emails covering top risers and decliners, and a public REST API for pulling the data programmatically. The tool requires no login and no registration, and no pricing is published on the site (it is presented as free). It is aimed at developers evaluating which open-source models are gaining traction, researchers tracking ecosystem dynamics, and teams deciding which Hugging Face models to adopt based on real growth signals rather than static popularity.

ML developers and researchers use ParaPulse to track growth momentum of open-source Hugging Face models and datasets to decide which ones are worth adopting.