Grab a Thumbnail Frame From Your Video

Your footage is a thumbnail source

YouTube offers three auto-generated frames after an upload, chosen by an algorithm with no idea what your video is about. They are almost always the wrong ones — a blink, a transition, the back of your head. Meanwhile the shot you actually want is sitting in the file on your disk.

This tool loads your video locally, gives you a scrubber and single-frame steps, and exports whatever frame you land on. Nothing is uploaded, so a 4 GB master works exactly as well as a compressed proxy and the file size is irrelevant.

Frame stepping

The ±frame buttons move in roughly 1/24 of a second — enough to walk past a blink or catch the peak of an expression. Browsers do not expose true frame boundaries, so this is a close approximation rather than an exact frame index; for finding a face at its best, it is more than precise enough.

Auto-grab six frames

Not sure where the shot is? Auto-grab seeks to six points spread evenly across the video, skipping the first and last 8% where intros and end cards live, and captures each one. It is a fast way to see what the video has to offer before you go hunting by hand.

Export sizes

1280×720 is the default: the frame is cropped to 16:9 and scaled to YouTube's thumbnail size, ready to upload. Choose full frame instead when you want the original resolution to work from — a 4K frame gives you room to crop into a face later.

If the video will not load

Playback here uses your browser's own decoders, so anything your browser cannot play, this cannot read. The usual culprit is HEVC (H.265), which many phones record by default and desktop Chrome often cannot decode. Re-export as H.264 MP4 and it will open.

Once you have your frame, put your words on it with the thumbnail text adder, or get it under the 2 MB limit with the image compressor.

Last updated August 15, 2026.