Speak, or bring a file.
Dictate and watch the text appear as you talk, or import MP3, M4A, WAV, AAC and FLAC. Silence is skipped automatically.

Offline engine ready
Record or import audio and get timestamped, editable text. It all runs on your phone, with no account and no upload.

The speech model runs on your phone, so recording, transcribing and editing never need a connection. Try the switch.
Dictate and watch the text appear as you talk, or import MP3, M4A, WAV, AAC and FLAC. Silence is skipped automatically.

Titles, dates and durations in one list, with full-text search across every transcript you have saved.

Tiny comes installed and works right away. Larger models take a one-time download and are more accurate.

Every word keeps its timestamp, even after you edit the words around it. Try it on this sample. It has no sound.
Thanks for meeting me. Before we start, this recording stays on my phone. That helps. Most of what I can share is about patient intake. Good. We can stop whenever you like.
Pick yours, or let the app detect it. Accuracy varies by language, voice and model, so try yours on the free models first.
Transcribe a confidential interview without handing the recording to anyone else's server.
Capture a whole lecture in a hall with no signal, then search it before the exam.
due process3 matches
Client and patient audio stays on the phone, encrypted with AES-256, behind an optional app lock.
Make subtitles for your videos locally and drop them into your editor.
1 00:00:04,000 --> 00:00:09,200 Before we start, this recording stays on my phone.
No subscription and no per-minute fees, because no server is doing the work.
The core app, for as long as you want it.
One-time purchase. Works offline after that.
No. Audio and transcripts stay on the device unless you export or share them yourself. The app only goes online when you download a larger model or buy or restore Pro.
No. There is no sign-up and nothing to log in to.
Recordings and transcripts are encrypted with AES-256 on your phone, with the keys held in its secure storage. You can also turn on an app lock.
There is no cloud copy, so your transcripts go with it. Export anything you want to keep somewhere else.
It depends on the language, the speaker, the recording and the model. Larger models are more accurate. Try your own audio on Tiny and Base, which are free.
EchoLocal has no HIPAA or SOC 2 certification. Keeping audio on the device can support your obligations, but check with your compliance team before using it for regulated work.
iPhones on iOS 16 or later and Android phones on Android 10 or later. If a model is too large for your phone's memory, the app marks it "Not recommended" and says why.