Upload a dog X-ray. Radiant’s AI runs in your browser, flags a possible fracture, boxes where it looked, and shows how confident it is.
¿Prefiere español? Todo Radiant está disponible en español: las páginas, los resultados del análisis y el informe PDF.
Radiant está en español: las páginas, los resultados y el informe PDF. Also available in English.
Real output from Radiant’s model on this X-ray. Hover the image or pick a view.
How it works
Three steps from an uploaded image to a highlighted region you can inspect, with the model's confidence and vote count shown alongside.
Upload a radiograph
A JPG, PNG or WebP of a dog X-ray, up to 20 MB. It’s analyzed right on your computer.
The model reviews it
It checks the image is an X-ray, then the model casts 10 votes on whether there’s a fracture.
Review the highlighted area
See the boxed area and heatmap, the model’s confidence, and save a PDF report.
Why we built Radiant
Dogs can't say where it hurts, so radiographs do much of the explaining. Radiant explores how an AI model can point a reviewer toward areas worth a closer look while staying open about its uncertainty.
Every result is framed as an area for additional review. Radiant never names a condition.
Model confidence appears next to every highlight, with a note on what it does and doesn't mean.
Radiant supports your review. Clinical decisions belong to qualified professionals.
Used in a real clinic
Radiant is in use at Alief Animal Hospital, our clinic partner. Their team uses it to review radiographs and says it works really well.
Questions
What Radiant does, what it doesn't, and how to read its output.