Transcriber Wiki

Quick start

1. Open Transcriber and allow microphone access if you want to record in the app.
2. Download a speech model from Models. The first download needs an internet connection.
3. Record audio or use Open file to import an existing recording.
4. Start transcription. The text appears as the recording is processed.

Speech models

Models are downloaded once and stored on the device. Smaller models need less storage and usually finish sooner; larger models can produce better results but use more CPU, memory, battery, and storage.

Set the spoken language when you know it. A fixed language is usually faster and more reliable than automatic detection.

Working with transcripts

Each transcript is kept with timestamped segments. Tap a segment while playing audio to seek to that moment. Use the library search to find a word or phrase across saved transcripts.

For difficult audio, try a larger model, set the correct language, reduce background noise before importing, and keep the phone connected to power during long jobs.

Export

Use the transcript actions to copy text or export it as TXT or SRT. SRT includes timestamps and is suitable for captions; TXT is better for notes and further editing.

Privacy and connectivity

Transcription is performed on the device after the selected model is available. The app does not require an account and does not upload recordings or transcripts to a transcription server. Internet access is used for model downloads and any external app you explicitly choose for sharing or translation.

See the full privacy policy for data handling and contact details.

Need help?

Email support@transcriber.40i.net with your Android version, device model, app version, and a description of what happened. Do not send private recordings unless support asks and you have permission to share them.