General
When enabled, current settings are applied to every loaded file during processing.
We cache the converter files on your browser so you don't have to re-download them every time, improving performance and reducing data usage.
Clear all site data including settings and cached files, resetting convert to its default state and reloading the page.
GPU acceleration is faster but can cause browser lag during AI tasks. CPU-only mode keeps the UI smooth.
Compress
Re-encodes video at best quality. Set a target size to constrain file size.
Image Tools
Get a free key at remove.bg/api. Images are sent to remove.bg servers.
Get a key at openrouter.ai/settings/keys. Prompts are sent directly to OpenRouter.
Text & Speech
TTS uses Kokoro 82M neural model (~92 MB). STT uses Whisper AI locally — model size depends on selection. Everything runs in-browser; no files leave your device.
Summarize
Larger models produce better summaries but take longer to download. Models are cached after first use.
Most websites block cross-origin requests. Enable this to route URL fetches through a third-party proxy (allorigins.win). Your requested URL will be sent to that server.
Everything runs in-browser; no files leave your device.
OCR
Tesseract.js WASM-based OCR. Language data is downloaded on first use (~15 MB per language). Everything runs in-browser; no files leave your device.
PDF Editor
Uses pdfjs-dist for rendering, Fabric.js for annotations, and pdf-lib for export. Everything runs in-browser; no files leave your device.