Files processed on compressimage.fun servers are automatically purged within four hours. You can delete completed jobs immediately. Browser-only tools never send your data to our servers.
1. Server-processed tools
Image compression, exact-size targeting, format conversion, resizing, cropping, rotation, metadata inspection/removal, watermark application, and favicon generation require native codec processing and upload files to our same-origin processing API.
Every job is isolated with cryptographic security:
- Random identification: Each job receives a cryptographically random directory ID and an independent 256-bit capability access token. Only a SHA-256 hash of the token is stored on the server.
- Storage isolation: Uploaded source files, encoded outputs, and private preview thumbnails are stored in an isolated, private volume under random internal filenames, not your original filename.
- Automatic TTL deletion: All job directories and files are permanently deleted by an automated cleanup daemon within four hours (14,400 seconds) of creation. Cleanup also runs automatically upon processor startup.
- Immediate manual deletion: Clicking "Delete now" in the tool workspace sends a
DELETEcommand that immediately and permanently removes the complete job directory and all associated files from disk.
2. Browser-only tools
The following developer and document utilities run 100% inside your local browser memory using client-side Web APIs (such as the HTML5 Canvas API, jsPDF, and typed arrays):
- Image to Base64
- Base64 to Image
- Base64 Image Viewer
- Image to Data URI
- Image to PDF
- Image Color Picker
When you use these tools, your images, pasted strings, and generated documents are never uploaded or transmitted to our servers. Network tests in our automated release gate strictly enforce this zero-network boundary.
3. Server logs and operational diagnostics
Our server operational logs are strictly limited to technical health and capacity monitoring. Logs may record:
- Request timestamp and internal request ID
- Tool identifier and operation mode (e.g.
compress:exact) - Input and output format (e.g.
jpeg,webp) - Coarse file size bucket (e.g.
under-1mb,1mb-5mb) - Processing duration in milliseconds and HTTP status code
- Controlled error category codes (e.g.
UNSUPPORTED_FORMAT,TARGET_IMPOSSIBLE)
Logs never contain image pixel data, Base64 strings, EXIF metadata, GPS coordinates, access tokens, job result URLs, or full user filenames.
4. Analytics and telemetry disclosure
We use analytics to monitor site reliability, understand which tools are popular, and improve performance across mobile and desktop devices.
Our analytics implementation operates under strict privacy safeguards:
- Google Analytics 4 (GA4): Configured with IP anonymization enforced and advertising features (Google Signals and ad personalization) explicitly disabled.
- Umami Analytics: A privacy-first, cookie-less analytics platform that does not track users across websites or collect personal identifiers.
- Semantic event boundaries: The event tracking layer passes only high-level operational categories (such as tool name, general format, coarse file count, and general success/error outcome). It never transmits filenames, image contents, EXIF tags, GPS values, job IDs, bearer tokens, or free-form error messages.
5. What metadata removal cannot do
Using our metadata removal tool strips EXIF tags, GPS coordinates, camera serial numbers, and embedded thumbnail records from image headers.
However, removing metadata does not blur or alter the visible pixels. If an image contains faces, house numbers, street signs, reflection details, or readable text on documents, that visual information remains present in the image.
6. Third-party sharing and monetization
We do not sell, rent, monetize, or share your uploaded files, metadata, or browsing activity with third-party advertisers, data brokers, or AI training registries.