Versely
    Guides

    Video Frame Extractor in the browser is not a generate

    /free-tools/video-frame-extractor never spends a credit. Do not pay a model to do an in-browser file job.

    Versely Team••4 min read

    /free-tools/video-frame-extractor never spends a credit. Do not pay a model to do an in-browser file job. Grabbing a still from a clip you already have is Canvas and a PNG. It is not image-to-video, not a thumbnail model, not an upscaler. If you are about to describe "export frame at 00:03" to a generator, stop. Open the extractor.

    Credits exist for jobs that invent pixels. The cost hub is the map of those jobs — per second, per character, per megapixel, per tier, quoted before you confirm. A frame grab is none of those meters. Treating it as a generate is how a team burns a video row to produce a screenshot.

    Grabbing a frame is a file job

    You already have the file. The pixels at that timestamp already exist. The work is: decode the frame, draw it at the video's own width and height, save a PNG. That is what the extractor does. It does not guess a prettier still. It does not restyle the product. It does not invent a thumbnail that "matches the vibe."

    Two uses show up constantly:

    • A thumbnail candidate from a take you already paid for — the splash, the hang, the expression between two poses.
    • A last frame to hand a model as the first frame of the next shot, so the next generate continues a character, a set, and the lighting instead of reinventing them.

    Both are file jobs. The second one feeds a generate. It is not a generate.

    What runs in the browser

    Load a video, scrub to the moment, export at the video's native resolution. No upload, no watermark, no account. The tool reads the file on your machine with the Canvas and File APIs, draws the current frame at videoWidth × videoHeight — not the on-screen player size — and offers a PNG. Lossless on the export; a heavily compressed clip still looks compressed.

    Nothing is sent to a server. It keeps working with the network off. That is why it is actually free, with no limit: the work happens on your device. Generation on Versely costs credits. There is no free tier for generation.

    Print-screen of the player grabs the compressed preview at layout size. The extractor is the source frame.

    Do not spend a generate on a screenshot

    The usual mis-buy: "make a still of this moment" into a model, with the clip as a vague reference. You pay a catalog row. You get an interpretation. The timestamp you wanted is gone. If you needed the true last frame of shot one as the conditioner for shot two, you just broke the chain.

    Another mis-buy: paying a thumbnail generator to recreate a frame you could have exported. Upscale later if the still is thin. Extract first. Then spend credits, if you still need to.

    The extractor will not invent a frame that was not in the file, fix resolution the source never had, or caption the still. It will give you the PNG. That is the whole product.

    Credits belong to jobs that invent pixels

    Every generation on Versely shows its credit cost up front. None of that arithmetic applies to a browser tool that never leaves the machine. If the next step is image-to-video from the PNG you just grabbed, that is a generate. Route it on a model page.

    If you needed a new picture, you needed an image model. If you needed the frame that is already in the clip, you needed the extractor.

    FAQ

    Does exporting a frame use credits?

    No. The video frame extractor is a free in-browser tool: no account, no limit, no upload. Credits apply to generation and to metered jobs on the cost hub, not to a PNG you draw locally.

    Is the PNG full quality?

    It is exported at the video's own pixel dimensions as a PNG, which is lossless, so the export does not degrade the frame beyond the source compression. A 720p clip yields a 720p still. That is honesty, not a defect in the tool.

    Why pull a frame instead of generating a thumbnail?

    Continuity. A still from the take matches the grade, the product, and the moment. A separately generated thumbnail is a second product. Pull candidates here; generate only if you need resolution, type, or a frame that never existed.

    Can I use the exported PNG as the first frame of the next shot?

    Yes. That is the second reason the tool exists. Export the true last clean frame and condition the next image-to-video job on it. Paying a model to imagine that frame is how identity resets at the cut.