Extract a frame from video — grab a still image (PNG or JPG) at any timecode. Great for thumbnails and stills, no watermark.
AI-powered audio review & delivery for voice production teams
Drop your video here or click to browse
MP4, MOV, M4V, WebM, MKV or AVI
Files are processed securely and auto-deleted — no account needed. See our Privacy and Terms.
One still frame, at the source resolution.
Want a short clip rather than a still? Make a looping GIF. Need to trim the video around the moment, or resize the source first? Those are separate tools.
Upload a video, pick a moment, and download the still as a PNG or JPG
Sometimes you do not need the clip — you need one perfect image out of it: the thumbnail for a YouTube upload, a still for a press kit, a product shot grabbed from a demo reel, or a reference frame to drop into a document or a review note. Scrubbing a video player to the right moment and taking a screenshot works, but it captures your screen — at screen resolution, with the player's controls, scaling and compression baked in. A frame extractor reads the video itself and writes out the actual decoded frame, at the video's own resolution, with nothing else in the picture.
This tool uses a technique called fast input-seek. Instead of decoding the whole video from the start until it reaches your timecode, ffmpeg jumps almost straight to the moment you asked for and decodes a single frame — then stops. Because it never processes the rest of the file, grabbing a frame from a 3-second clip and grabbing one from a two-hour recording take about the same (very short) amount of time. There is no re-encode of the video and nothing is written back to it: the tool only ever reads one frame and saves that as an image. That single-frame decode is the entire job, which is why the size and length limits here are far more generous than on tools that have to transcode a whole video.
You can save the frame as a lossless PNG or a compressed JPG. The right choice depends on what the still is for:
| PNG | JPG | |
|---|---|---|
| Quality | Lossless — pixel-for-pixel the decoded frame | Slight compression, usually invisible |
| File size | Larger | Smaller |
| Best for | Thumbnails, screenshots, graphics, crisp text and edges | Photographic frames, quick shares |
| Re-editing | Ideal — no generation loss | Fine, but each re-save loses a little |
If you plan to edit the still afterwards, add text, or you want the sharpest possible thumbnail, choose PNG — it is the default because it preserves exactly what was in the video. If you just need a lightweight image to send or post and the frame is photographic, JPG keeps the file small.
Whatever resolution your video is, that is the resolution of the extracted frame. A 4K (3840×2160) video yields a 4K still; a 1080p video yields a 1920×1080 still. The tool does not upscale, downscale or crop the frame — if you need a different size or aspect ratio, resize the source first or crop the exported image afterwards.
Because only one frame is decoded, the caps are generous: without an account you can use a file up to 200 MB and 30 minutes; a free account raises that to 500 MB and 120 minutes. You extract one frame per run — to grab several moments, just run it again at each timecode. Your upload is processed securely, is never shared, and both the upload and the exported image are deleted automatically shortly after processing — no account is required to use the tool.
Upload your video, type the moment you want as a timecode (for example 0:30), choose PNG or JPG, and click Extract Frame. The tool decodes that single frame and gives you an image to download — no screen-recording or screenshotting needed, and the still comes out at the video's own resolution.
Enter the time in the Frame time field as MM:SS or HH:MM:SS — for example 0:30 for thirty seconds in, or 1:02:15 for just past an hour. 0:00 grabs the very first frame. If you enter a time at or past the end of the video, the tool tells you how long the video actually is so you can pick a valid moment.
Choose PNG for the crispest result — it is lossless, so the still is pixel-for-pixel the decoded frame, which is ideal for thumbnails, graphics, text and any image you will edit further. Choose JPG when you want a smaller file for a quick share or email and the frame is photographic; the compression is usually invisible. PNG is the default.
The frame is saved at the source resolution of your video. A 4K video gives a 4K still, a 1080p video gives a 1920×1080 still, and so on — the tool does not upscale, downscale or crop it. If you need a different size or aspect ratio, resize the source before extracting or crop the exported image afterwards.
Right now the tool grabs one frame per run. To capture several moments — for a contact sheet or to compare shots — just run it again at each timecode; each extraction is quick because only that single frame is decoded. Batch and every-N-seconds export are on the roadmap as a future option.
MP4, MOV, M4V, WebM, MKV and AVI. The tool checks the actual streams in the file rather than trusting the extension, so an audio file or a mislabelled upload is rejected with a clear message. The output is always an image — a PNG or a JPG — regardless of the input container.
VoiceDeck adds AI-powered audio & video review and delivery for your whole team — so every file ships in spec, automatically.
