Updated October 2026
Grabbing one frame at a time works fine for a single still, but if you need several stills from the same clip — a storyboard, a set of thumbnail candidates, or a handful of reference images pulled from a longer video — repeating scrub-step-grab over and over is slow. Batch export automates that: it steps to several evenly-spaced points across the whole video in one pass and hands you a grid to choose from.
Set the "Number of frames" field (anywhere from 2 to 30) and click "Grab evenly-spaced frames." The tool divides your video's total duration into that many equal intervals and seeks to each one in turn, drawing every resulting frame onto the canvas and capturing it — so a 10-frame batch on a 100-second video grabs roughly one frame every 10 seconds, start to finish. It's a systematic sample across the whole timeline, not a random or content-aware selection — the tool has no idea what's happening in the footage, just where in time each grab lands.
Each grabbed frame appears as a thumbnail in a grid below the batch controls, already selected by default. Click "Select none" and pick just the ones you want, or "Select all" to keep everything — useful when you asked for more frames than you need just to see more options, then narrow it down visually once you can actually see the thumbnails side by side.
Once you've settled on a selection, "Download selected as ZIP" bundles them into a single archive using the open-source JSZip library, built entirely in your browser — the frames never pass through a server to get packaged, the same way nothing else in this tool touches a server. Each file in the ZIP is named with its sequence number and timestamp, so you can tell at a glance which point in the video each still came from without opening every file.
A higher frame count is useful when you're not sure exactly which moment you want and would rather scan a denser set of options than scrub manually — for example picking the best possible thumbnail from a talking-head clip, where small timing differences change facial expression a lot. A lower count (2–6) is enough for a quick storyboard overview of a whole video's structure, one still roughly per scene or section.
It won't detect scene changes, faces, or "interesting" moments — it's purely time-based, evenly spaced across the duration you loaded. If you need one very specific moment rather than an even sample, use the single-grab workflow with frame-by-frame stepping instead; see the finding an exact frame guide for that. You can also run batch export first to get oriented, then switch back to manual stepping near whichever batch thumbnail turned out closest to what you actually wanted, to fine-tune from there.
Batch export seeks and decodes each frame individually, the same as a manual grab does — so a 30-frame batch involves 30 real seek-and-decode operations, one after another. On longer or higher-resolution videos this takes noticeably longer than a single grab, but it still runs entirely on your own device with no upload, and there's no server-side queue or size limit beyond your device's own memory and patience.
Is there a limit on how many frames I can batch-grab? The input accepts 2 to 30 frames per batch run.
Do batch-grabbed frames have the same resolution and watermark limits as a single grab? Yes — the same free-tier 960px cap and watermark apply to every frame, whether grabbed individually or in a batch.
Can I re-run batch export with a different count on the same video? Yes — change the number field and click the button again; it re-samples the loaded video at the new interval without needing to re-upload anything.
Does the ZIP include anything besides the images? No — just the selected frame images, named by sequence and timestamp.