ToolGPT

Grab a still frame from a video as a PNG or JPEG

Need a thumbnail, a cover image or a screenshot from a recording? Attach the video to ToolGPT and ask for a still. It shows six moments with their timecodes; tap one and it is saved as an image in your browser.

  • Runs in your browser
  • No sign-in needed
Drop or choose your filesThe assistant opens with your files and does the rest.Choose Files
Things you can type
product_demo_final.mp4
Open ToolGPTAttach a file and describe what you need.

What it can do

  • Six moments to choose from

    Stills spread across the whole video, each labeled with its timecode.

  • Zoom in on a moment

    Say “around 1:30” and ToolGPT shows six frames from the two and a half seconds around it.

  • Straight to an exact time

    Name a timestamp and that frame is saved without a gallery.

  • PNG or JPEG

    PNG by default, with no quality loss; JPEG when you want a smaller file.

  • Full size, right way up

    The still has the video's own display size, so a portrait phone video gives an upright portrait image.

  • Ready for the next step

    Compress the still, convert it, or put several into a PDF in the same chat.

How it works

  1. 1

    Attach the video

    Add it with the + button or drop it into the chat. No account is needed.

  2. 2

    Pick the moment

    Ask for a still. Tap one of the six frames, ask for frames around a time, or name the exact second.

  3. 3

    Download the image

    The still appears as an image card with a download button. The video is not changed.

Finding the right frame without a scrubber

The first six frames are spread evenly through the video, stepping past the very first and last moments, where black leaders and end cards usually sit. If the shot you want falls between two of them, say “a bit later” or give a rough time, and ToolGPT shows six frames packed into a short window around it.

The time reported for the image is the time of the frame actually saved. Video is a series of separate frames, so a request for 12 seconds can land on a frame that starts a fraction of a second earlier, and ToolGPT tells you which one you got.

One picture at full quality

Each request saves one still image at the video's full resolution, so a 4K video gives a 4K picture. PNG keeps every detail and suits documents and uploads. Ask for JPEG when the file needs to be smaller, which matters most for 4K stills.

The frame is decoded in your browser, so the video is not uploaded. An AVI or FLV file has to be converted to MP4 first, which ToolGPT can do in the same chat. If you want movement rather than a still, ask for a GIF instead.

Frequently asked questions

Can I save every frame of a video?

No. The frame grabber saves one still per request. For a short moving clip, make a GIF.

What resolution is the image?

The same as the video's picture. A 1080p video gives a 1920×1080 still.

Is my video uploaded to grab a frame?

No. The still is taken from the video in your browser.

Try it with your own file

Open ToolGPT, attach a file, and type what you want done.

Open ToolGPT

Conversions

Related tools

See all tools