Pro-level ball flight. No hardware.

They still cost thousands, still need their own case, and still have to be set up before you can hit a ball. Meanwhile the phone in your pocket ships with a high-speed camera and a depth sensor — most of the hardware problem, already solved, already paid for.

The remaining problem was software, and it was not a small one. Reading spin off a golf ball at 240 fps and inferring what the club did to produce it is genuinely hard, and it had to run on the phone rather than in a data center — partly for privacy, mostly because a round trip to a server costs more time than the whole computation does. That took a long time to get right.

Accuracy claims are cheap, so we tested ours in the open. In live side-by-side testing against a professional-grade launch monitor, LiPAR matched ball speed, carry and spin to within 2%. Not in a lab — head to head, shot after shot, on camera. When a number here is soft, the app tells you: the LiPAR Index carries a confidence band that stays wide until the model has seen enough of your swing to narrow it.

The models learned from real golfers before that was possible. Our beta users hit the shots that trained them, in their own garages and bays, against their own nets — which is why the app works in the places people actually practice, rather than only in the places a demo gets filmed.

Engineered on-device. Verified on the range.