Sintesi is an offline speech to text app. It turns your voice into text you can actually use — and it does it on your phone, not on someone else's server.
Record voice notes, meetings, interviews or ideas — or share a voice message you received. Sintesi transcribes the audio on the device, then rewrites it in whatever form you need: a clean note, a bullet list, meeting minutes, a summary, or a message ready to send. Every version is kept, and you can edit any of them by hand.
OFFLINE TRANSCRIPTION, ON-DEVICE AI
The AI models run entirely on your phone. There is no account, no sign-up, no subscription and no server of ours.
Your recordings are never uploaded, because there is nowhere for them to go.
A voice message is not a neutral thing to hand to a server. It carries a voice, a name, a place, a mood. Your recordings, transcriptions and generated text stay in the app's private storage, and are deleted with it.
This is private AI designed for your own recordings: the processing happens on your device, not in a data centre.
WHAT IT DOES
• Voice recorder with a live level meter and waveform
• Offline speech to text. Convert audio to text and voice to text without a network connection.
• AI note taker. Rewrite the same transcript as a clean note, list, meeting minutes, summary or ready-to-send message.
• Meeting transcription and meeting notes from a single recording, including summaries, discussion points, decisions and action items.
• Offline translation of your notes. Each language pack is a one-time download of about 30 MB.
• Receive voice messages shared from WhatsApp and other apps.
• Import audio and video files already stored on your phone.
• Tag, search, archive and restore notes from a 30-day bin.
• Export everything whenever you want.
• Light and dark themes, in English, Italian, German, French and Spanish.
WHO IT IS FOR
Journalists, students, researchers, and anyone who records interviews, lectures, meetings or voice memos and prefers not to send audio to a cloud service.
Sintesi is a transcription app for people who want their recordings and notes to remain under their control.
BEFORE YOU INSTALL
• Sintesi downloads about 3 GB of AI models during setup, preferably over Wi-Fi. Those models are what make offline transcription and AI processing possible. You choose which ones to install.
• AI rewriting requires a phone with at least 8 GB of RAM. On devices with less memory, recording and transcription continue to work, but rewriting features are unavailable.
• Transcription takes minutes, not seconds. The work happens on your phone instead of in a remote data centre.
• Imported audio can be up to 30 minutes long. Recordings created inside Sintesi have no length limit.
• The app is free and displays a banner ad on some screens.
• Requires Android 8.0 or later.
PRIVACY
Sintesi contains no analytics, no telemetry and no crash reporting. We do not know how often you open the app, or what you do with it.
The network is used for only two things: downloading AI models and serving the banner ad. To display ads, Google receives technical information about your device: the IP address, and an advertising identifier only if you agree to personalised ads.
The content of your notes never leaves your phone: not the audio, not the transcriptions, not the generated text.
Record, transcribe, summarise and translate audio with on-device AI. Your voice stays on your phone.
Full privacy policy: https://skaonline.github.io/sintesi/privacy.html
Offline speech to text. AI transcription, summaries, notes. No cloud or account.