ALL VIDEO TOOLS23
Video Thumbnail
Video Thumbnail Maker — Free Frame Grabber
Scrub to a moment, capture the frame the decoder presents there, and download it as an image.
Nothing is enforced, but the clip is decoded by this tab's own player and the frame is drawn into a canvas at full size, so what binds is the memory a browser gives one page rather than anything measured about the file.
About Video Thumbnail
A poster frame, a social card, a still to put in a document — one picture out of a video, taken from a moment you choose. Load the clip, scrub the player to the frame you want, press Use the player's position or move the time slider, and capture. The picture comes off your browser's own video decoder through a canvas; there is no conversion engine to download and the file is never uploaded. Two numbers on the page are reported rather than assumed, and both matter. The first is the time. A video carries pictures only at its own frame times, so a seek lands the playhead where you asked and the decoder presents the first frame at or after it — which is normally within a frame of your request and is occasionally further. What you get back is stamped with the time it really came from, and that is the time in the filename. A request inside the last quarter-second of a clip is also pulled back, because the frame walk needs a moment after the seek to present anything, and the page warns about that before you click rather than leaving you to notice it in the filename afterwards. The second is the size. The width choice never upscales: asking a 640-wide clip for a 1280-wide still gives you 640, so the picture is always real pixels, and the readouts print the size that was actually written. JPEG is the usual choice and the smallest; PNG is lossless and much larger, worth it when the still is going to be edited rather than posted; WebP is smaller than JPEG at the same setting, with a PNG fallback and a note on screen in browsers that cannot write it from a canvas. One frame is one frame — this is a still from your video, not a re-encode of it, so the picture quality is whatever the source carries at that moment.
Questions
Why is the captured time different from the one I asked for?
A video carries pictures only at its own frame times. The seek lands the playhead where you asked and the decoder presents the first frame at or after it, so the still can come from slightly later. The result is stamped with the time it really came from and that is the time in the filename — the page prints both numbers rather than only the one you typed.
Why can't I capture right at the end of the clip?
The frame walk needs a moment after the seek for the decoder to present something, so a request inside the last quarter-second is pulled back. The page warns about that before you click and names the time it will use instead.
Can I get a bigger image than the video?
No. The width choice never upscales — asking a 640-wide clip for a 1280-wide still gives you 640 — so the picture is always real pixels rather than an interpolated guess. The readouts print the size that was actually written.
Which format should I use for a poster frame?
JPEG, usually: small and universally supported. PNG is lossless and much larger, worth it when the still is going to be edited rather than posted. WebP is smaller than JPEG at the same setting, with a PNG fallback and a note on screen in browsers that cannot write it from a canvas.
Does capturing a frame re-encode my video?
No — the video is not written at all. One frame is drawn onto a canvas and encoded as an image, so the picture quality is whatever the source carries at that moment. No conversion engine is downloaded and nothing is uploaded.
Is my file uploaded to a server?
No. Transmute processes everything locally in your browser using JavaScript and WebAssembly. Your files never leave your device — there is no server, no upload, no cloud processing.