Textracto

AlexPechkin

About this app

Textracto is a powerful, privacy-centric tool that turns speech and images into text directly on your device. No cloud, no data collection, and no subscription required — just pure local AI performance.

Key Features:
• Speech-to-Text with Whisper or Parakeet: Transcribe any audio — voice notes, meetings, lectures — using local speech recognition models, fully on-device.
• Share-to-Transcribe: Send audio files (MP3, WAV, M4A, OGG/OPUS) from any app — Telegram voice notes, WhatsApp, file managers — and transcribe them in seconds.
• Background Transcription: Recognition keeps running in the background with progress notifications, even with the screen off. Queue multiple files while you continue with your day.
• Image-to-Text (OCR): Extract text from photos and screenshots with local Tesseract OCR. Crop the area you need, pick recognition languages, and get editable text. English is bundled; more language packs available in-app.
• Textracto Keyboard: Dictate or snap a photo directly from your keyboard — recognized text is inserted right into the field you're typing in, in any app.
• Floating Dictation Button: A floating voice recognition button that appears whenever the on-screen keyboard is shown. Tap it, speak, and the recognized text is automatically inserted into the text field you are currently typing in — in any app.
• Your Personal Hub: All recordings and images with their recognized texts are stored in one place. Play audio with word-by-word highlighting, edit, rename, or share results anytime.
• Model Store: Browse and download developer-recommended recognition models directly in the app.

Why choose Textracto?
• No API keys required after model setup — no internet connection needed for recognition.
• Full control over your AI models and data.
• Perfect for students, journalists, and anyone converting speech and images into text.

The Floating Dictation Button feature uses Android's Accessibility Service (AccessibilityService API). The service is used only for the following purposes:
- to detect when the on-screen keyboard appears, so the floating button can be shown and hidden accordingly;
- to detect the currently focused text input field;
- to insert the recognized text into that field.

Textracto does not use the Accessibility Service to read, collect, store, or share any screen content or personal data, and does not transmit anything off your device. All speech recognition is performed locally on your device. The service is entirely optional: it is disabled by default, requires your explicit permission in Android Settings > Accessibility, and you can turn it off at any time. All other features of the app work without it.

Important Note: The app does not include pre-installed speech recognition models. Use the built-in Model Store to download recommended models. English OCR support is bundled; other language packs are available in the Model Store.

What’s new

Meet Floating Dictation Button - works with any keyboard!