Dashboard
Extract text from anything.
Images, audio, video, PDFs and documents — recognized, transcribed and analyzed right in your browser. Nothing is uploaded.
Recent activity
Image OCR
Drop an image to recognize every visible character. Choose the document language for the best accuracy.
Audio Transcription
Transcribe speech to timestamped text using a Whisper model that runs locally. The model downloads once on first use.
Video Text Extraction
Sample frames for on-screen text and transcribe spoken audio, then merge everything into one searchable result.
Document OCR
Open Word, OpenDocument, rich text, HTML, Markdown and plain text files to read and analyze their contents.
PDF Reader
Read native PDF text instantly, or OCR scanned pages. Navigate page by page and export the result.
Batch Processing
Queue many files of any supported type. Process them in parallel with pause and cancel controls.
History
Every extraction is logged on this device. Search, reopen, export or clear it at any time.
Settings
Personalize the interface and processing. Everything is stored locally on this device.
Appearance
Processing
Data & history
Xikvak AI Extractor
A private, browser-native platform for turning files into editable, searchable text — no servers, no uploads.
Privacy by design
All recognition, transcription and analysis happen entirely inside your browser using WebAssembly. Your files never leave this device, and your history lives only in local storage.