Skip to main content
FRAME GRABBER · NOTHING LEAVES YOUR DEVICE

Grab a still frame from any video

Load a video, scrub to the moment you want, and save that single frame at full resolution. It happens in this tab — the file is never uploaded, and there is no account to make.

FREE · NO UPLOAD · PNG OR JPEG AT SOURCE RESOLUTION

Frame grabber

Choose a video file, move to the frame you want, and capture it. Everything runs in your browser.

The file is read locally and drawn to a canvas. It is not uploaded, and no copy is kept.

HOW IT WORKS

Four steps, all of them local

Nothing here talks to a server. The video is read from your disk, drawn once to a canvas, and handed back as an image file.

  1. 01

    Choose the video

    Any file your browser can play. Large files are fine — they are read from disk rather than copied or uploaded.

  2. 02

    Find the frame

    Drag the scrubber for the rough position, then step forward or back one frame at a time to land exactly.

  3. 03

    Capture it

    The frame is drawn at the video’s own resolution, so a 4K source gives a 4K still. Save as PNG to keep every pixel, or JPEG for a smaller file.

  4. 04

    Use it

    A still is the most useful thing you can own: a thumbnail, a reference, or the start frame of an AI clip.

WHY THIS ONE

Frame-accurate stills, and the footage never uploads

Your video is read in this tab and stepped through frame by frame, so the still you save is the one you chose, at the video’s own size.

It never uploads

Your video is read through a local object URL and drawn to a canvas in this tab. There is no request, no queue and no copy on a server to delete later.

It steps, not guesses

A screenshot catches whatever was on screen. Stepping frame by frame at a chosen rate lets you land on the one frame where the expression is right.

It keeps the resolution

The frame is captured at the video’s own dimensions rather than at the size of the player. A 4K source gives you a 4K still, not a window-sized one.

PAST THE PAGE

A still is where a good AI clip starts

Generating video from a text prompt asks the model to invent the whole scene. Starting from a frame fixes the composition, the colour and the subject before generation begins, which is why it usually takes fewer attempts to get something usable.

  • Give the frame to the Video Studio as a start frame and describe only the motion — one camera move, one action.
  • Draft at 360p for 30 credits, then re-render the take you want at 1080p for 125.
  • In many countries a new account gets 100 free starter credits, no card. Unused starter credits expire 72 hours after they are granted.

This page is free and needs no account. Generating video is a billed mode; the pricing page lists what a clip costs.

See how image-to-video works

QUESTIONS

Frame grabbing, answered

Is the frame extractor free?

Yes, and there is no account. It runs entirely in your browser, so it costs us nothing to serve and costs you nothing to use. Generating video from the frame afterwards is a separate, billed thing.

Does my video get uploaded?

No. The file is read from your disk through a local object URL, drawn to a canvas, and turned into an image — all inside this tab. Nothing is sent anywhere, so there is no copy on a server for anyone to keep.

Should I save a PNG or a JPEG?

PNG if the frame is going anywhere else — it is lossless, so nothing is thrown away. JPEG when you want a smaller file and the frame is going straight into something that will re-compress it anyway.

Is the stepping frame-exact?

It is as exact as a browser allows. Browsers do not expose a video’s real frame rate, so stepping moves by one frame at the rate you choose. If a step lands between two frames, nudge it again — the scrubber is accurate to a millisecond.

Can I use the frame to make an AI video?

That is the main reason to grab one. A start frame fixes the composition before generation begins, so the model only has to handle motion, which usually takes fewer attempts than describing a whole scene from nothing.

Can I animate a frame with a real person in it?

No. Image-to-video refuses raw photographs of identifiable real people, deliberately — that is a consent boundary. Frames of products, places, animals, artwork and generated characters all work. A real person appears through Storyboard Studio, which they set up and control.

FROM A FRAME TO A CLIP

You have the frame.Now make it move.

A still you chose beats a scene the model invented. Bring the frame to the Video Studio, describe the motion, and draft it cheaply before you commit.