Privapixel
Official documentation — Images, audio, and video on your computer
What is Privapixel?
Privapixel is a Windows application that combines practical tools for images, audio, and video in one place: compress, resize, remove background, apply watermarks, round corners, convert formats, enhance image, prepare ready-for-sale photos, create offer art, transcribe audio, transcribe video, download videos by link, and generate viral clips from long videos. Processing runs locally on your PC — your files are not uploaded to external servers.
- Compress — Reduce file size with adjustable quality, output format, and optional metadata removal.
- Resize — Change dimensions in pixels or by percentage, with aspect ratio lock and no-upscale option.
- Remove background — AI mode (U²-Net via ONNX) or fast mode for solid-color backgrounds.
- Watermark — Combine logo, text, phone, website, Instagram, and QR code with position and opacity controls.
- Rounded corners — Apply corner radius with live preview and optional transparent PNG export.
- Convert formats — Transform images between major formats with automatic source detection, quality control, and optional metadata removal.
- Enhance image — Adjust sharpness, noise, contrast, brightness, and colors; fix white balance and saturation; export with Save for Web for smaller files.
- Ready for Sale Photo — Turn casual product shots into professional marketplace images: local AI background removal, centered product, white or premium background, soft shadow, and marketplace-ready export sizes.
- Offer Art — Create promotional art with product photo, prices, highlight line, WhatsApp button, and optional store logo — ready for WhatsApp, Instagram, and social media.
- Transcribe Audio — Convert MP3, OGG, OPUS, M4A, AAC, WAV, and other audio files to text (.txt) or subtitles (.SRT) with local Whisper AI.
- Transcribe Video — Extract and transcribe audio from MP4, MKV, AVI, MOV, and other videos, exporting text or .SRT subtitles for video editors.
- Download Video — Save videos to your computer from links, choose quality and format, download audio only, organize downloads in a queue, and optionally export subtitles.
- Viral Clips — Send a long video and local AI finds the best moments, generates short clips with viral potential scoring, synced subtitles, and word-by-word highlight — in vertical (9:16), horizontal (16:9), or the original format.
- Video Light — Batch-compress videos with H.264 and constant quality (CRF), with ready-made profiles, processing queue, and thumbnails in the queue and preview panel.
- File queue — Add multiple files, preview original vs. result, and batch process.
- Drag and drop — Import images, audio, video, and folders directly into the drop zone.
- Output folder — Choose where to save results and open the folder with one click.
- Windows with .NET Framework 4.8 (or a compatible version indicated by the installer).
- Enough disk space for the app, AI models (when used), and processed images.
- Internet is required only to activate the license, validate registration, check for updates, and on first transcription to download the Whisper model (~142 MB) — image, audio, and video processing does not depend on the cloud.
- For AI background removal, the ONNX model may be downloaded to the app data folder on first use.
- For audio and video transcription, FFmpeg is required (bundled with the installer or available in the app folder). Formats such as MP3, OGG, and OPUS are converted internally before transcription.
- For the Download Video tab, the installer includes the yt-dlp engine and FFmpeg. On some sites, detection uses the Microsoft WebView2 Runtime (usually already present on Windows 10/11).
- For the Viral Clips tab, FFmpeg is required (included in the installer or available in the program folder) and the local Whisper model (~142 MB), downloaded on first transcription.
- For the Video Light tab, FFmpeg is required (included in the installer or available in the program folder).
- You receive the Privapixel installer. Source code is not included in the package.
- After purchase, use your email and purchase code in the download and license registration flow when available on the website.
- Run the installer, complete setup, and open Privapixel from the Start menu or shortcut.
- On first launch, complete license registration with the details sent by email.
- The app may check for updates automatically on startup by comparing your version with the official server.
Before using the application, you must register your license on the activation screen.
- Enter the email used for purchase.
- Enter the license code you received.
- Registration ties the license to your computer name (machine identifier).
- After successful validation, the app opens normally. Data is stored in
%AppData%\Roaming\Privapixel\reg.dat.
Each license is valid for one computer. Contact support if you need to move to a new machine.
After activating your license, the main interface has fourteen tabs: Compress, Resize, Remove background, Watermark, Rounded corners, Convert formats, Enhance image, Ready for Sale Photo, Offer Art, Transcribe Audio, Transcribe Video, Download Video, Viral Clips, and Video Light.
- Add files — Use the add button or drag images, audio, video, and folders into the drop area on the left.
- Queue and thumbnails — Select an item to preview the original and the result (transcription on the audio and video tabs; suggested clips on the Viral Clips tab; video frame on the Video Light tab).
- Bottom panel — Configure options for the active tab and set the output folder.
- Process — The main action button (Compress, Resize, etc.) applies the operation to the queue or selected item.
- Open folder — Opens the output folder in Explorer after processing.
- Clear — Empties the queue to start a new batch.
The app remembers your last import folder to speed up repeated work.
Reduce file weight for web, email, or storage with fine control over quality and format.
- Mode — Choose Maximum quality, Balanced, or Smallest size.
- Quality — Adjust the percentage slider (especially for JPEG and WebP).
- Output format — Automatic, JPEG, PNG, or WebP.
- Strip metadata — Removes EXIF and similar data to save space and improve privacy.
- Only if smaller — Keeps the original file when compression would not reduce size.
The side preview shows size estimates and lets you compare visually before saving in batch.
Change width and height to standardize catalogs, thumbnails, or social posts.
- Pixels — Enter desired width and height. With Keep aspect ratio, changing one value recalculates the other.
- Percentage — Use shortcuts for 75%, 50%, or 25% of the original size.
- Do not upscale — Prevents small images from being stretched beyond their original dimensions.
Set the output folder and process the queue. Resized images are saved in the chosen folder with a name similar to the original.
Isolate the main subject and produce a file with transparency (PNG).
- Local AI (U²-Net) — Recommended for photos, products, and portraits. The model runs on your PC via ONNX; output is PNG with a transparent background.
- Fast mode (solid color) — For flat backgrounds (white, green, etc.). Adjust tolerance (fuzz), automatic color detection, and edge smoothing.
On first use of AI mode, the model may be downloaded. This requires disk space and internet only for the initial download.
Protect or brand your images by combining visual and text elements.
- Logo — Select an image file as your mark.
- Text — Custom phrase or name.
- Phone, Website, and Instagram — Optional contact fields.
- QR Code — Generated from the text or URL you provide.
- Position — Bottom right/left, top right/left, center, or custom with X and Y coordinates.
- Opacity — Control how transparent the watermark appears over the image.
Enable only the elements you need; the preview shows the result before batch apply.
Round image corners in batch — ideal for avatars, product cards, thumbnails, and social media assets.
- Corner radius — Slider from 0% to 50%, with pixel value shown alongside and live preview.
- Transparent background (PNG) — Export with transparency in rounded corners; preview uses a checkerboard when transparent areas are present.
- Side-by-side preview — Compare the original with the result before processing the queue.
- Batch processing — Apply the same radius to dozens or hundreds of images at once.
Set the output folder and process the queue. Final files are saved in the chosen folder with a name similar to the original image.
Improve photo quality in batch — especially useful for phone pictures, catalogs, and social media. Select one or more options and combine adjustments as needed. Everything is processed locally on your PC.
- Increase sharpness — Enhances edges and detail with controlled sharpening.
- Reduce noise — Smooths grain common in low-light or small-sensor photos.
- Improve contrast — Adds more separation between light and dark areas.
- Improve brightness — Slightly increases overall brightness.
- Correct colors — Balances levels and tones for more natural color.
- Brighten dark image — Recovers underexposed or very dark photos.
- Correct white balance — Reduces yellow or blue color casts (indoor light, flash, shade).
- Boost saturation — Makes colors a bit more vivid, ideal for washed-out photos.
- Save for Web — Optimizes for online publishing: resizes to up to 1920 px on the longest side, smart compression (JPEG, WebP, or PNG), and strips metadata. Appears on the same row as the output folder path.
Select at least one enhancement (or only Save for Web). The side-by-side preview shows the estimated result before you click Enhance image. Saved files use the _melhorada suffix in the output folder.
Turn casual product photos into professional images for Mercado Livre, Shopee, catalogs, and online stores — in one click, with 100% local processing on your PC.
- Local AI background removal (U²-Net) — The model runs on your computer; your images are not uploaded to external servers.
- Cutout and centering — The product is isolated, refined, and positioned automatically in the composition.
- White background (marketplace) — Ideal for platforms that require a clean, neutral background.
- Premium background (studio) — More sophisticated, studio-style look.
- Soft shadow — Optional gentle shadow under the product for depth.
- Preset sizes — 1000×1000 (Mercado Livre), 1200×1200 (recommended), 1600×1600 (high quality), and 2000×2000 (premium).
- Save for Web — Option to further optimize the output JPEG.
- Real-time preview — Compare the original with the estimated result before processing the queue.
Set the output folder and click Prepare photo for sale. Saved files use the _pronta_venda suffix in JPEG format.
Create promotional art automatically for WhatsApp, Instagram, and social media — ideal for sellers who post offers with price, highlight text, and a call to action.
- Product name — Title shown on the promotional band.
- Old and new price — Previous price appears struck through; promotional price is highlighted.
- Highlight line — Optional urgency phrase (e.g. Limited-time offer).
- CTA (WhatsApp) — Button text; can include a phone number or phrase such as Call on WhatsApp.
- Use WhatsApp logo — Checkbox to show or hide the official WhatsApp icon before the text on the button.
- Store logo — Optional; appears discreetly in the corner of the art.
- Automatic layout — Product photo highlighted at the top and a red promotional band at the bottom with a white rounded button.
- Instant preview — See the estimated art while filling in the fields; ideal for adjusting copy before batch generation.
Click Generate offer art to save JPEG files to the output folder with the _arte_oferta suffix. Enter at least the price or product name to generate the art.
Convert audio recordings into readable text or synchronized subtitles using local Whisper speech recognition — ideal for voice messages, podcasts, meetings, and audio files exported from messaging apps.
- Input formats — MP3, OGG, OPUS, M4A, AAC, WAV, AMR, FLAC, WMA.
- Language — Portuguese, Auto, English, or Spanish.
- Output — Text (.txt) for reading and copying, or Subtitles (.srt) with timestamps for use as captions.
- Batch queue — Add multiple audio files and transcribe them all with one click.
- Preview — View transcribed text on screen before and after saving to the output folder.
- Remove from queue — Right-click the item and choose Remove audio.
On first transcription, the Whisper model (ggml-base.bin, ~142 MB) is downloaded to the app data folder. Compressed formats (MP3, OGG, OPUS, etc.) are converted internally to WAV before transcription.
Extract audio from a video and transcribe it automatically — perfect for Reels, TikTok, YouTube, recorded lessons, and testimonials. Generate text or ready-to-import .SRT subtitles for your video editor.
- Input formats — MP4, MKV, AVI, MOV, WEBM, WMV, FLV, M4V.
- Automatic extraction — The app extracts audio from the video and sends it to local Whisper with no cloud upload.
- Language — Portuguese, Auto, English, or Spanish.
- Output — Text (.txt) or Subtitles (.srt) with start and end time for each line.
- Subtitles for editors — Import the .SRT into CapCut, Premiere, DaVinci Resolve, and other editors without typing manually.
- Batch queue — Transcribe multiple videos at once.
- Remove from queue — Right-click the item and choose Remove video.
Processing is 100% local — your videos are not uploaded to external servers.
Save videos to your computer from links — ideal for archiving your own content, authorized material, or public-domain media. Paste the address, choose quality and format, and organize downloads in a queue, all processed locally on your PC.
- Supported links — YouTube, Vimeo, TikTok, and hundreds of other sites supported by the built-in download engine.
- Add link — Paste or type the URL and click Add link, or press Enter. With Smart mode enabled (default), pasted links are added to the queue automatically.
- Type — Choose Video or Audio only.
- Quality — Best available, 4K (2160p), 1440p, 1080p, 720p, 480p, or 360p (for video).
- Format — MP4, MKV, or WebM for video; MP3, M4A, or OGG for audio.
- Subtitles (PT/EN) — Optional for video downloads when available from the source.
- Playlists — If the link is a playlist, the app asks whether to download all videos or only the link you entered.
- Download queue — Add multiple links, select an item to view details, and click Download to start.
- Output folder — Choose where to save files and open the folder with one click when finished.
The first time you click Download, you must accept the terms specific to this feature. Use only with your own videos, authorized content, public-domain material, Creative Commons works, or when you have permission from the rights holder — you are fully responsible for the processed link and for how you use the downloaded file.
Turn long videos into short clips ready for social media — ideal for podcasts, lessons, live streams, interviews, and testimonials. Local AI analyzes the content, suggests the best segments, ranks each clip by viral potential, and exports selected clips with optional subtitles, all processed locally on your PC.
- Input formats — MP4, MKV, AVI, MOV, WEBM, WMV, FLV, M4V.
- Add videos — Use Choose videos... or drag files and folders into the drop zone.
- Language — Portuguese, Auto, English, or Spanish (for transcription and subtitles).
- Clip duration — Short (15–35s), Medium (25–60s), or Long (45–90s).
- Max clips — 5, 8, 10, or 12 clips per video.
- Export format — Vertical 9:16 (Reels, TikTok, Shorts), Horizontal 16:9, or Original (keeps the source video aspect ratio).
- Subtitles — Check Include subtitles on exported clips to burn synced captions into the final files.
- Word-by-word highlight — With subtitles enabled, optionally highlight the spoken word in real time (karaoke style), one phrase at a time.
- Turbo mode — For long videos, transcribes only promising segments (~5× faster). Recommended for lengthy content.
- Suggested clips — After analysis, the right panel lists each cut with score, viral potential, duration, segment, and reason for the suggestion, ranked from most to least promising.
- Output — MP4 files in the output folder.
- Remove from queue — Right-click the item and choose Remove video.
Click Generate viral clips to analyze the selected video (or the entire queue) and export clips to the output folder.
Processing is 100% local — your videos are not uploaded to external servers. Long videos may take several minutes to analyze and export, depending on duration, clip count, and selected options.
Batch-compress videos with constant quality (H.264 / CRF) — reduce file size while keeping good visual quality, with 100% local processing on your PC.
- Input formats — MP4, MKV, AVI, MOV, WEBM, WMV, FLV, M4V.
- Add videos — Use Choose videos... or drag files and folders into the drop zone.
- Thumbnails — When loading, the queue shows a frame from the video (extracted locally via FFmpeg). When you select an item, the preview shows the original on the left; after a successful compression, the compressed file thumbnail appears on the right.
- Profile — Balanced 1080p (recommended), Fast 1080p, High quality 1080p, Fast 720p, Keep original resolution, or Custom (adjust CRF quality).
- Quality (CRF) — Available in the custom profile; lower values preserve more quality and higher values reduce file size more.
- Web optimized — Option for MP4 with fast start for web playback.
- Only save if smaller — Keeps the original when compression does not reduce file size.
- Queue and details — Columns for Original, Compressed, and Status; the bottom panel shows resolution, duration, profile, and savings after processing.
Click Compress video to process the queue. Compressed files are saved to the chosen output folder.
Long or high-resolution videos may take several minutes to convert, depending on profile and hardware.
On startup, the app checks the official server for a newer version. If one is available, you get a prompt to download the installer and review release notes.
We recommend keeping the latest version installed for fixes, new formats, and performance improvements.
Version 1.7.0.0
- Video Light — New tab to batch-compress videos with H.264 and constant quality (CRF): ready-made profiles for 1080p and 720p, custom CRF, web optimized option, save only if smaller, and a queue with video thumbnails in the list and preview panel when loading files.
- Isolated crop (Ready for Sale Photo) — Export only the product as transparent PNG, without background.
- Logo on Viral Clips — Apply your logo to exported clips, with configurable position and opacity.
Version 1.6.0.0
- Viral Clips — New tab to turn long videos into short clips: local AI finds the best moments, ranks by viral potential, exports in vertical (9:16), horizontal (16:9), or original format, with synced subtitles, word-by-word highlight, and turbo mode for long videos.
Version 1.5.0.0
- Download Video — New tab to save videos to your computer from links: choose quality, format (video or audio), optional subtitles, download queue, playlist support, and smart mode when pasting links.
Version 1.4.0.0
- Transcribe Audio — New tab to convert audio to text with local Whisper AI; choose language, export .txt or .SRT, and process multiple files in a queue.
- Transcribe Video — New tab to extract audio from video and transcribe automatically; export text or .SRT subtitles for CapCut, Premiere, DaVinci Resolve, and other editors.
Version 1.3.0.0
- Ready for Sale Photo — New tab for sellers: local AI background removal, centered product, white or premium background, optional soft shadow, and JPEG export at 1000×1000, 1200×1200, 1600×1600, or 2000×2000 pixels.
- Offer Art — New tab to create promotional art with name, prices, highlight line, WhatsApp CTA, optional WhatsApp icon, and store logo; side-by-side preview and batch processing.
- Support covers installation, licensing, and use of the features described in this documentation.
- Custom software development beyond the product scope is not included.
- Custom work may be quoted separately.
- Each license is valid for one computer.