98 %. And we’re very specific about what that means.

When a lab technician uploads a model to an AI platform and receives a result that still needs some tweaking—a mesh artifact here, a poorly processed ring there—the instinct is often to correct it manually and move on to the next case. In many labs, this workflow has become routine: generate, check, correct, repeat.

But this correction process eventually takes its toll. And above all, it reveals a deeper problem: if your technicians have to go back and fix the model behind the AI, then the AI isn’t really doing the work. They’re the ones doing it.

This is where the concept of “success rate” needs to be clarified. For many solutions, a case is considered successful as long as the scan is accepted and a model is generated.

But a generated model isn’t necessarily ready for printing. If you still need to repair the mesh, clean up certain elements, or adjust the file before sending it to production, then the work isn’t done. And in many cases, the model simply isn’t ready yet.

This isn’t a success. It’s just the first step.

EasyModel is based on a much more rigorous standard. For us, a successful scan means that the scan is accepted and that the resulting model is truly ready to be printed: no retouching, no manual cleanup—it’s ready to go straight into the production queue.

EasyModel meets this standard in more than 98% of cases.

For labs that have incorporated retouching as a standard step in their workflow, this difference is crucial. What you may consider routine today is not set in stone. That time can be reclaimed and reinvested elsewhere.

EasyFlow goes even further. Its on-device AI doesn’t just process cases—it gradually learns from the actual changes made by your technicians, allowing it to adapt more and more precisely to your lab’s standards. The goal isn’t simply to provide an acceptable starting point. The goal is to produce the correct result on the first try, with ever-increasing reliability, specifically for your laboratory.

There’s a real difference between AI that assists your teams and AI that creates more work for them. A true success rate must measure results that can be put to immediate use. And that’s the only number that really matters.