Extract every frame from a video at the source frame rate — right in your browser. Export PNG, JPG, or WebP in a ZIP. Free, no upload, no watermark.
Timestamps are best-effort in this browser; each exported filename records the exact captured timecode.
No install, no upload, no account. Everything happens in this tab.
Drop an MP4, MOV, or WebM file. The browser reads it locally and shows duration, resolution, and frame rate.
Pick timestamps, an interval, a frame count, or every frame. The frame math updates automatically.
Review the thumbnails, keep only the frames you need, and download a ZIP with numbered filenames and a manifest.
Many online extractors cap output at 30 fps or around 1920px. This tool reads the video's own frame rate and keeps the original resolution.
The detected frame rate is shown next to duration and resolution, so you can confirm what every-frame mode will capture.
Frames are captured from the decoded video at its native width and height. Nothing is resized for upload.
If a browser cannot decode a file, the tool says so. It does not swap in a lower-quality preview.
Step through the video frame by frame, or type timecodes to capture only the moments you need.
Enter several timestamps at once, or move one frame at a time with the step buttons.
Every file includes its frame number and captured timecode, so a folder of stills stays sortable.
The ZIP ships with a manifest.json that lists the source video, settings, and each frame's timestamp.
Pick the format that fits the next step: editing, publishing, or feeding an AI workflow.
Lossless frames with the largest files. Best for editing, text, UI, and anything that will be re-exported.
Adjustable quality and the smallest files. A practical choice for thumbnails, previews, and quick sharing.
Small files with good quality and broad browser support. A good default for modern sites and apps.
AI clips are short. Pull the final frame, use it as the first frame of the next image-to-video run, and keep the shot moving.
Choose First and last, then download both stills in one ZIP. The final frame becomes the starting image for the next clip, while the first frame is useful for continuity checks.
The frame never needs to leave your device for this step. When you are ready to generate, upload it to the tool below and describe the motion you want next.
There is no upload step, no server-side queue, and no account. The file is opened directly by the browser.
The video is read from your device and drawn to a canvas in the same tab. No API receives the file.
Frames live in memory until you download them. Closing the tab clears the session.
Exports are clean and there is no account to create. The limits are stated on this page.
A frame is often the fastest route to the next asset. These are the workflows people bring to this tool.
Grab a sharp, well-lit moment and use it as a thumbnail, cover, or social post without screenshotting a player.
Pull key beats from a clip at an interval, then arrange the stills into a storyboard or moodboard.
Export the first or last frame, then use it as a reference image or the next image-to-video start frame.
Step through a short clip to compare poses, check continuity, or document how an effect changes over time.
Different tools solve different parts of the problem. This table compares the facts that matter for frame extraction.
| This tool | Ezgif | 123apps | FFmpeg | |
|---|---|---|---|---|
| Runs in the browser | Yes | Yes, with upload | Yes, with upload | No |
| Original resolution | Yes | Capped around 1920px | Limits depend on the tool | Yes |
| Every frame at source fps | Yes, within browser limits | Up to 30 fps | Varies by tool | Yes |
| No watermark or signup | Yes | Yes, with ads | No | Yes |
| Batch ZIP with manifest | Yes | Yes | Yes | Scripted |
| Setup required | None | None | None | Install and commands |
Resolution, frame rate, privacy, and limits — answered without marketing language.
Frames are captured at the video's native width and height. Every-frame mode uses the detected frame rate instead of a fixed 30 fps cap. If the rate is variable, the tool reports that instead of guessing.
Both. Choose Every frame for a complete sequence, or use Timestamps, Every N seconds, Every N frames, or Evenly spaced to capture a shortlist. First and last captures the two frames used most often in AI video workflows.
No. The file is read by your browser and decoded locally. There is no upload endpoint, no server-side queue, and no account, so the video never leaves your device.
Exports are watermark-free and free to use. The practical limit is the browser's memory: a single run is capped at 1,000 frames and 1.5 GB per file, which is enough for short clips and typical social video.
MP4 (H.264), MOV, WebM, and other formats the browser can decode. If a codec is unsupported, the tool tells you and suggests converting the file first; it does not pretend to process it.
The tool reports the detected frame rate and records the captured timecode for every file. Browser seeking is best-effort, so a frame may land on the nearest decodable frame; the manifest lets you verify exactly what was saved.
The page loads on mobile browsers, but large jobs depend on device memory. On iOS, start with single frames or small batches and expect slower seeks than on desktop.
Yes. Export the last frame of a clip, then use it as the first frame of an image-to-video run to continue the shot. ReelMuse's image-to-video tool is built for that workflow.
Take the frame you just saved and give it motion. Upload it to Photo to Video, describe the shot, and generate a clip.