A clearer view of hail. See the evidence.

Silent recorded weather film. Three synchronized views show Anvil’s hail signal, MESH radar hail sizes, and ProbSevere storm objects over Southeast Denver metro, Colorado on August 13, 2026. All three systems show this storm. This replay compares their visual presentation; it does not establish comparative accuracy. 25 recorded scans cover 21:42:39-22:30:41 UTC. Contours morph between recorded scans. Consecutive recorded scans. Intermediate shapes are display interpolation, not observations. Fixed maps preserve each system’s geometry. Display opacity is enhanced for Anvil.
About this recording
All three systems show this storm. This replay compares their visual presentation; it does not establish comparative accuracy. ProbSevere v3 and MESH are named for comparison. No affiliation or endorsement is implied. The linked ground-report cases carry the detection evidence.
25 recorded scans · 21:42:39-22:30:41 UTC. Original colors and geography are preserved. Anvil opacity is enhanced for display. Contours morph between recorded scans. Consecutive recorded scans. Intermediate shapes are display interpolation, not observations.
Recording & sources© OpenStreetMap contributorsMore storms. Ground reports.
3 selected casesHail detected. ProbSevere’s hail score stayed below the comparison threshold.
1″ reported · estimated1 report · View case →SizeNorman
Estimate within the reported range.
1-1.25″ reported · estimated3 reports · View case →ContextCapon Springs
ProbSevere flagged this storm too.
1.75″ reported · estimated1 report · View case →
The app that knows your address. Not your county. Your roof.
Real-time hail alerts for the address you care about: how big, when it arrives, and what to do, with live radar and your neighborhood’s hail history. Free on iPhone and Android.
FREE TIER · NO CREDIT CARD · TAKE-COVER ALERTS NEVER PAYWALLED
Free vs Premium, and everything in the appPRODUCT REPLICA · SAMPLE HOME · FORECAST · INSIGHTS · RADAR
Every storm, verified to the address. A console for the teams who work the aftermath.
Roofing, restoration, and claims teams run their storm desk in Hail Sentinel: territories and pinned properties, live radar and replay, alert rules, address-level verification reports, and post-storm lead lists, with an API underneath. The beta is by approval.
PRODUCT REPLICA · DEMO WORKSPACE · RADAR · LEADS · CLAIMS
-
Territories & properties
Draw the areas you monitor or canvass and pin the properties you protect. One map for the whole operation.
-
Radar & storm replay
Live radar with storm tracks over your territories. Replay a storm over an address and export it as video.
-
Alert rules
Rules decide who is told, when, and how. Each rule keeps its own delivery history.
-
Verify reports
Confirm hail at an address and generate claim-ready PDF or CSV reports, one-off or on a schedule.
-
Leads & canvassing
Post-storm homeowner leads in a full pipeline, and door lists built from confirmed storms, ranked and routed to crews.
-
API & workspace
REST API with issued keys and signed webhooks. Team roles, an assistant that answers questions across your data, and an audit log.
From raw atmosphere to a single clear answer.
Four data sources collapse into one hail decision, delivered to the app on your phone and the console on your desk.
Four data streams, continuously.
Radar, satellite, ground reports, and physics models are fused in real time across the contiguous United States.
Identify the hail signature.
A dual-pol ML pipeline sifts the noise and identifies hail cores, size, and track.
Two ways out of one record.
Alerts to the app on your phone, and a console and API for the teams who work the aftermath, in beta by approval.
The app
Free on iPhone and Android. Hail size, arrival, and what to do about it, for every address you watch.
The business console
Territories, storm replay, address-level verification reports, and a REST API. The beta is by approval.
The whole Lower 48, every two minutes.
Radar across the contiguous United States, processed at 1 km resolution on 2-minute update cycles.