FAQ
What the tools read, what they write, and what happens to your files. The short answer to the last one: nothing leaves your device.
Every toolEvery tool in ImagesCompress ImageResize ImageConvert ImagePhoto Collage MakerBlur FacesCrop ImageRemove Image MetadataMarkdown to PDFExtract PDF pagesQR Code GeneratorMute VideoExtract AudioCompress VideoTrim VideoVideo to FramesVideo to GIF
Every tool
- Are my files uploaded?
- No. Every tool runs entirely in your browser and nothing is sent to any server. The production Content-Security-Policy stops the page from opening a connection to anywhere except the analytics endpoint, so this is enforced by your browser rather than promised by us.
Every tool in Images
- Which formats can I open?
- JPEG, PNG, WebP, GIF, BMP, ICO and AVIF work in every modern browser, and so does HEIC: Safari reads it through the system codecs, and everywhere else the page loads a bundled decoder the first time a HEIC arrives — which is why that first file takes a moment longer. TIFF, JPEG XL and JPEG 2000 are the ones that still depend on the browser: Safari reads them, Chrome and Firefox mostly do not. A file yours cannot read is skipped, and the others still go through.
Compress Image
Compress JPEG, PNG, WebP, GIF, AVIF and HEIC images without uploading them anywhere. Set the quality, keep the original format or convert to JPEG, PNG or WebP.
- Can I compress a PNG?
- Yes. PNG is lossless, so saving a PNG back as PNG ignores the quality slider — to make it smaller pick WebP, which keeps transparency, or JPEG when the image has none.
- What format do I get back?
- “Same as input” keeps the original format whenever your browser can write it, and otherwise picks the closest one it can: a HEIC photo becomes JPEG, a GIF or BMP becomes PNG. You can also force JPEG, PNG or WebP.
- What happens to animated GIFs or WebP?
- Only the first frame is processed — the result is a still image.
Resize Image
Resize JPEG, PNG, WebP, GIF, AVIF and HEIC images locally in your browser. Set exact pixels or a percentage, optionally keeping the aspect ratio, and keep the original format or convert to JPEG, PNG or WebP.
- Can I keep the aspect ratio?
- Yes — enable “Keep aspect ratio” and only the width is used to scale proportionally.
- What format do I get back?
- The one you dropped in, whenever your browser can write it — otherwise the closest it can manage: a HEIC photo comes back as JPEG, a GIF or BMP as PNG. Forcing JPEG, PNG or WebP is also an option, and it changes only the format, never the size you set.
- Is transparency preserved?
- Yes, as long as the output format has an alpha channel — PNG and WebP do. Saving to JPEG flattens transparent areas onto white.
Convert Image
Convert HEIC, JPEG, PNG, WebP, GIF and AVIF images without uploading them anywhere. Drop a batch, pick the format you want, and download the results.
- Can I convert HEIC from my iPhone?
- Yes, in any modern browser. Safari reads HEIC through the system codecs; everywhere else this page loads a decoder the first time a HEIC arrives, which is why that first file takes a moment longer.
- What happens to transparency?
- PNG and WebP keep it. JPEG has no transparent pixels, so converting to JPEG places the image on a white background.
- Why can’t I pick the format my files are already in?
- Because that would not convert anything. In a mixed batch you can pick it, and the files already in that format are copied through untouched rather than re-encoded.
- Can I choose the quality?
- Not here — conversion uses a high quality setting so the result matches the original. To make files smaller, use Compress Image.
- What about animated GIFs or live photos?
- Only the first frame is converted — the result is a still image.
Photo Collage Maker
Build a photo collage entirely in your browser: choose a grid of 2, 3, 4, 6 or 8 frames, a square, portrait, story or landscape canvas, and adjust the border and corners. Drag inside any frame to set what shows.
- How many photos can I use?
- Between 2 and 8 in the collage itself. You can load more than that — the extras wait in the tray and you can swap them in at any time.
- Can I change the grid without starting over?
- Yes. Photos stay where they are when you switch layout. If the new grid has fewer frames the extra photos go back to the tray rather than being lost.
- Can I choose which part of a photo shows?
- Yes. Tap a frame to select it, then drag inside it to move the framing and use the zoom slider to get closer.
- What size is the finished collage?
- You choose: 1080, 2048 or 4096 pixels on the long side. The shape follows the canvas you picked.
- Can I remove the border?
- Yes — set the border to 0 and the photos meet edge to edge. You can also change its colour, which is the same colour as the background.
Blur Faces
Blur faces in a photo entirely in your browser. The tool detects faces automatically; you move, resize, remove or add boxes, choose blur, pixelate or a solid fill, and download — with nothing ever uploaded.
- What if it misses a face?
- Draw a box over it by hand. Automatic detection is never perfect — profiles, small or partly hidden faces can be missed, so you are always the final check.
- Can I choose how to cover a face?
- Yes: blur, pixelate (mosaic) or a solid fill, each with an adjustable strength, per box or applied to all at once.
- Is the blur reversible?
- A light blur or mosaic can sometimes be partly reconstructed by software. The solid fill is the only fully irreversible option.
- Does it need a connection?
- Only to load the page. Once loaded, everything works offline — the detector runs on your device.
Crop Image
Crop an image entirely in your browser. The selection starts as the whole photo: drag its edges or handles to the area you want to keep — freely, or locked to 1:1, 4:3, 3:2, 16:9 or the photo’s own ratio — watch the exact pixel size as you go, and download. Nothing is uploaded, and the file is rewritten from the cropped pixels alone.
- Can I crop to an exact ratio?
- Yes. Pick 1:1, 4:3, 3:2, 16:9 or Original and both the drag and the handles keep it; Free is the default. Choosing a ratio refits the selection you already have rather than starting over.
- What format do I get back?
- The one you dropped in, whenever your browser can write it — otherwise the closest it can manage: a HEIC photo comes back as JPEG, a GIF or BMP as PNG. Forcing JPEG, PNG or WebP is also an option, and it changes only the format, never the area you selected.
- Does it keep the photo’s metadata?
- No. The file is written from the cropped pixels, so EXIF — including GPS — does not survive. Orientation does: it is applied to the pixels before you pick the area, so what you select is what you get.
- Can I crop without a mouse?
- Yes. Tab to the image, then the arrow keys move the selection a pixel at a time, and Shift with an arrow key resizes it — left and right move its right edge, up and down its bottom edge — keeping any ratio you locked.
Remove Image Metadata
Inspect what a photo’s EXIF metadata reveals — camera model, date and time, GPS location when present, and author or rights tags — then remove it entirely in your browser. Cleanup is lossless wherever the format allows; otherwise the image is re-exported without its metadata, and the result always says which happened.
- What’s the difference between lossless and re-export cleanup?
- Lossless rewrites the file to drop metadata segments without touching a single image byte. When that isn’t possible for a format, the image is decoded and re-encoded instead — which can change quality, size or even format — and the result always states which one happened.
- Does opening the GPS link send my photo anywhere?
- No. It opens OpenStreetMap with only the latitude and longitude in the URL — never the file or the report — and only when you click it.
- What’s in the batch ZIP?
- The cleaned copies plus a report.json listing every file’s source name, outcome, strategy and warnings, so nothing is silently dropped or renamed.
- What if a file can’t be cleaned?
- It fails on its own row with a plain-language reason; the rest of the batch still processes and downloads.
Markdown to PDF
Convert a Markdown document to PDF without uploading it anywhere. Type or paste your Markdown, or open a .md file, and a live preview shows what the file will contain: headings, lists, tables, task lists, highlighted code, LaTeX formulas and Mermaid diagrams. Then download a PDF made right in your browser — no print dialog, no headers added by the browser, selectable text and clickable links — while the document never leaves your device.
- Where is the PDF made?
- In your browser. The page builds the file itself: there is no print dialog, nothing is uploaded, and the browser adds nothing — no date or web address at the top of each page.
- Why are some of my images missing?
- An image on the web would have to be fetched, and this page makes no requests with your document — so it becomes a labelled box. Images embedded in the Markdown as data URIs are kept in the preview, and in the PDF when they are PNG or JPEG; other formats appear as their description. Images stored next to a .md file on your disk cannot be reached from the file alone.
- Which Markdown does it understand?
- GitHub-flavoured Markdown — tables, task lists, strikethrough and autolinks — plus highlighted code blocks, LaTeX formulas between $ signs, Mermaid diagrams in mermaid code blocks, and plain HTML tags, which are cleaned before they are shown.
- How do formulas come out in the PDF?
- A formula on its own line, between $$ signs, is drawn as a sharp vector picture. A formula inside a sentence keeps its LaTeX source in the PDF, because a line of PDF text cannot hold a drawing — put it on its own line if it should be drawn.
- Can I choose the paper size?
- Yes — A4 or US Letter, next to the download button. It starts on A4. The margins are fixed.
- Why are some characters missing from my PDF?
- The PDF carries its own fonts. Body text is set in Roboto, which covers Latin, Greek and Cyrillic; arrows, check marks and box-drawing characters are drawn from the code font, DejaVu Sans Mono. Emoji, Chinese, Japanese, Korean and right-to-left scripts such as Arabic or Hebrew are in neither, so the tool warns you before you download.
Extract PDF pages
Extract pages from a PDF without uploading it. Choose a file, tick the pages you need, and download a new PDF containing just those pages in their original order. The work happens in your browser, so the document and everything in it stay on your device.
- Does the order of my clicks change the result?
- No. The new PDF always follows the page order of the original, whatever order you ticked the boxes in.
- Can I extract pages from a password-protected PDF?
- Not here. A protected file is refused with a message instead of being copied into a result that would not open. Remove the password in a PDF reader first, then try again.
- Is the original PDF changed?
- No. The tool reads the file and writes a new PDF containing only the selected pages; the file on your device is left exactly as it was.
- What happens to links and form fields?
- The pages are copied as they are, so their text, images and layout come through. Interactive form fields that span several pages may stop working in the new document, because only the selected pages are copied.
QR Code Generator
Generate a QR code from a link or any line of text without uploading it anywhere. The code is drawn by your own browser, so nothing reaches a server — and because it contains your link directly rather than a redirect, it does not expire and it counts no scans. Choose the size, the colours, the quiet zone and the error correction level, then download it as a PNG or as a scalable SVG.
- Does the QR code expire?
- No. The code contains your link or text directly, so there is no short URL in the middle and no redirect through anyone’s server. It keeps working for as long as the link it points to does, and nobody can count the scans.
- Should I download the PNG or the SVG?
- The SVG for anything that gets printed: it is a vector, so it stays sharp at poster size. The PNG for everything else — a slide, a message, a web page.
- What does error correction do?
- It adds redundancy, so a code that is scratched, wet or partly covered still reads. Higher levels recover more but pack more modules into the same square, which makes each one smaller.
- Can I use any two colours?
- You can, but scanners need contrast, and a light code on a dark background defeats many of them outright. The tool warns you before you download rather than after you print.
- Why does my code get denser as I type?
- A QR code grows in fixed steps. Longer text, or a higher error correction level, means a bigger grid in the same square.
Mute Video
Remove the audio track from a video without uploading it anywhere. The encoded picture is copied into a fresh container untouched, so there is no generation loss and no waiting for a re-encode.
- Does this re-encode my video?
- No. The encoded video is copied straight into a new container, so the picture is bit-identical to the original and the job takes seconds rather than minutes. Only the audio track is left out.
- What file do I get back?
- An MP4 for H.264 and HEVC video, and a WebM for VP8, VP9 and AV1. Those three can technically sit in an MP4, but very few players will open one, so they go where they play. A .mov comes back as .mp4, which every player opens.
- Why is the file smaller?
- Because the audio track is gone. The video itself is unchanged, so the saving is whatever the soundtrack weighed.
Extract Audio
Take the audio track out of a video without uploading it anywhere. The track is copied into an audio container untouched, so there is no generation loss — and no choice of output format, because choosing one would mean re-encoding.
- Is the audio re-encoded?
- No. The track is copied out of the video and into an audio container untouched, so it is bit-identical to what was in the file. That also means you get the format the video already held, not one you choose.
- What file do I get back?
- An .m4a for AAC, MP3 and FLAC audio, and a .webm for Opus and Vorbis. An .m4a holding Opus is something almost nothing plays, so those go where they work. Most phone and camera video holds AAC, so .m4a is the usual answer.
- Can I get an MP3?
- Not from this tool, because converting to MP3 means re-encoding, which loses quality and is a different job. This tool only ever copies.
- It says my video has no audio track.
- Then it genuinely has none — a silent screen recording, or a clip whose sound was already removed. There is nothing to extract.
Compress Video
Compress MP4, MOV, WebM and MKV videos without uploading them anywhere. Pick a resolution and a quality; the audio track is compressed too when this browser can, or copied through otherwise.
- What happens to the audio?
- It is compressed alongside the video where this browser supports it, and copied through unchanged otherwise. On a browser that cannot encode audio in the format this produces, the track may be dropped instead — a case Mute Video also covers on purpose.
- Can this make my video bigger by mistake?
- It can, on a video already compressed harder than the quality you pick here — the same way re-saving an already-small JPEG at high quality can grow it. The size shown next to the result tells you either way.
- Why does a long video take a while?
- Every frame is decoded and re-encoded, unlike Mute Video or Extract Audio, which only rewrite the container. A video past three minutes gets a note saying so before the heavy part starts.
Trim Video
Cut a piece out of an MP4, MOV, WebM or MKV video without uploading it anywhere. Drag the two handles to the part you want to keep and download it: the cut lands exactly where you put it, and trimming only the end copies the video through untouched.
- Is the cut exact, or does it jump to the nearest keyframe?
- Exact. The clip starts and ends where you put the handles, to the frame. Tools that cut on keyframes can move your cut by a second or two without telling you, which is why this one does not work that way.
- Why is cutting the end instant while cutting the start takes minutes?
- Cutting only the end lets us copy the video across untouched, so nothing is re-encoded and the result is identical to the original. A cut that starts later cannot be copied — every frame from your start point on has to be decoded and encoded again, which is the same work Compress Video does.
- Does trimming lose quality?
- Not when you only move the end handle: those bytes are copied, not re-encoded. Moving the start handle re-encodes the picture once, at a high quality setting. The audio is re-encoded on that path too, and on a browser with no encoder for it the track may be dropped — Mute Video covers that case on purpose.
Video to Frames
Extract still images from a video without uploading it anywhere. Drag the handles to the part you want, choose a frame every few seconds or every single frame, and download the lot as one ZIP. The stills come out at the video’s own resolution, and the whole thing runs in your browser.
- What size and format are the frames?
- Full size — the same resolution the video plays at, with any rotation already applied, so a clip shot on a phone comes out the right way up. You choose JPEG, PNG or WebP: JPEG is the default most people want, WebP usually comes out a little lighter still, and PNG keeps every pixel exactly, at several times the size of either. If your browser cannot write the format you picked, the frames are saved as PNG and named .png, rather than given an extension that would be a lie.
- How many frames can I extract at once, and why is there a limit?
- The count is shown while you choose, and it warns you once it is too high. Everything happens in your browser, so every frame and the ZIP around it live in your device’s memory at the same time — and how many fit depends on the format you picked and the video’s own resolution: a lossless PNG costs several times what a JPEG or WebP frame does, and a 4K frame costs four times what a 1080p one does. A shorter piece of the video or a longer gap between frames brings the count down directly. Switching from PNG to JPEG or WebP doesn’t change the count — it raises the limit, since PNG costs several times more per frame than either.
- Why is the count approximate when I choose "every frame"?
- Because many videos do not run at a steady frame rate — phone recordings and screen captures often vary — and the real number is only known once the file has been read. We measure the rate from a sample of the video and mark the count with a ~ when it is an estimate. When the file says its rate is constant, the number is exact and shown without one.
Video to GIF
Make an animated GIF out of part of a video, without uploading it anywhere. Drag the handles to the piece you want, pick how wide the result should be and how many frames a second, and the GIF is built in your browser — then you can watch it before you download it.
- Why is there no 15 frames per second?
- Because a GIF cannot store it. Every frame carries how long to wait before the next one, as a whole number of hundredths of a second — so 10 frames per second is exactly 10 hundredths, 20 is 5, and 25 is 4. Fifteen would be 6.67, which has to be rounded to 7, and the file would then play at 14.29 frames per second rather than the 15 you picked. Other converters offer it and quietly write 7. We would rather only offer the rates that come out the way they read.
- Why does the length I can pick change with the size?
- A GIF frame costs about one byte per pixel, so a 960-pixel-wide frame costs four times a 480-pixel one, and the whole animation is held in your device’s memory while it is built — nothing is uploaded, so there is nowhere else for it to be. The count you can take is shown while you choose, and it goes up the moment you pick a smaller size or fewer frames per second. At the original size of a full-HD video you get about two seconds, which is the honest reading: a GIF is not a video codec.
- Why is the GIF bigger than the video it came from?
- Because GIF is a format from 1989 that stores every frame almost independently, with at most 256 colours each, while an MP4 stores mostly the differences between frames using thirty years of newer compression. Growing is the expected cost of the format you asked for, not a sign anything went wrong — which is why this tool does not show a "smaller/larger" badge on the result. Picking a smaller size is the one lever that changes it a lot.