Recorded test
See where the model would click
Choose an example and compare its click with the target.
Live OCR and text matching; a simpler alternative to the GUI-Actor model used in the recordings.
Recorded examples
The red dot is the model’s click. The green box marks the right target. These are saved test results, not a live model.
Loading recorded predictions...
Exact position and full-size image
View the full-size screenshotMark a point on your own screenshot
This is a manual tool. Your image stays in this browser; no model runs on it.
Click the image to mark a target point, or focus it and use Enter and the arrow keys. Hold Shift for larger steps. The saved point is measured against the original image.
No point selected.