Extract Video Frame

Capture the exact frame you need

Videos often contain a still image that is useful on its own: a product shot, a thumbnail, a slide from a presentation, a frame from a tutorial or a scene you want to reference in a report. This tool extracts one frame from a local video at the timestamp you enter and saves it as a PNG or JPG image. You can type a simple number of seconds, a decimal value for finer control or a mm:ss timestamp that is easier to copy from a video player.

The extraction happens in your browser. Instead of uploading the whole video to a remote service, the page loads ffmpeg.wasm, writes the selected file into the in-memory media engine and asks ffmpeg to seek to the requested time. The generated image is then downloaded through the browser, and temporary files are removed from the ffmpeg file system.

PNG for screenshots, JPG for smaller files

PNG is the default because it keeps sharp edges, text and interface details clean. It is a strong choice for software demos, slides, charts, captions or any frame that should not show compression artifacts. JPG is usually smaller and works well for camera footage, scenery, people and other photographic frames. If you plan to edit the image further, start with PNG when size is not a concern; if you need a lightweight preview, JPG is often enough.

Timing and privacy notes

Frame extraction depends on the source video's timestamps, keyframes and codec. If the image is a little before or after the moment you expected, adjust the timestamp by a fraction of a second and run the extraction again. Very large or high-resolution videos can take longer because all work is done on your device. The first use may pause while the WebAssembly engine downloads, but after that the same tab can reuse the loaded engine for additional captures without another full setup.