Memotori Support
Setup, frequently asked questions, and how to reach us.
Contact
For bug reports, questions, and requests, please write to web-admin@s1products.info.
Including the following makes it easier to find the cause:
your macOS version (Apple menu > About This Mac) / your Memotori version (Memotori menu > About Memotori) / your Mac (Apple Silicon or Intel) / the transcription model you are using / what you did, and what happened
There is no need to send us your recordings or transcripts.
Getting started
1. Download a transcription model
Transcription requires a speech model. On first launch, open “Transcription Models” in the sidebar and download one. Until a model is present, transcription produces no output.
| Model | Best for |
|---|---|
| tiny / base | Trying it out. Fast, but less accurate |
| small / medium | Balancing speed against accuracy |
| large-v3 turbo | Meetings and anything where accuracy matters (recommended) |
Models are retrieved from Hugging Face (huggingface.co), and are downloaded only when you ask for them. Models you no longer need can be deleted from the same screen.
2. Add audio
Drag and drop audio or video files onto the window. Multiple files are processed in order, and you can also send files from Finder’s “Open With”.
3. Summarize
Once the transcript is ready, run a summary to get an overview, decisions, and to-dos. The language model used for summarization is bundled with the app, so there is nothing else to install.
Permissions
| Permission | Used for |
|---|---|
| Microphone | Recording and dictation |
| Screen Recording | Recording the audio your Mac is playing (for example remote participants in a call). This is how macOS gates system audio; Memotori does not capture, store, or transmit the contents of your screen |
| Accessibility | Inserting dictated text at the cursor in other apps |
You do not need to grant a permission for a feature you do not use. Permissions can be revoked at any time in System Settings > Privacy & Security.
Frequently asked questions
Transcription produces nothing
Most likely no model has been downloaded. Open “Transcription Models” in the sidebar and check that one is installed.
The first transcription after launch is very slow
The first run after launching optimizes the model for the Neural Engine, which can take a few minutes. Subsequent items are faster. Choosing a smaller model helps when you are in a hurry.
The other participants’ voices are not recorded
In Google Meet, Zoom, or Teams, their voices come out of your speakers and never reach your microphone. In the recording screen, choose “this Mac’s audio” or “microphone and this Mac’s audio”. No virtual audio device such as BlackHole is needed.
If it still does not work, check that Memotori is enabled in System Settings > Privacy & Security > Screen Recording. You may need to restart Memotori after granting it.
Dictated text is not inserted into other apps
Enable Memotori in System Settings > Privacy & Security > Accessibility. Without it, the text is copied to the clipboard instead, so you can paste it with ⌘V.
A word is always transcribed incorrectly
Add the incorrect and correct spellings to “Term dictionary” in the sidebar, and it will be corrected automatically from then on. This works well for names of people and companies.
Speaker names are not filled in automatically
A person is remembered by company, name, and voice characteristics once you have named them, and is filled in automatically next time. Assign a name in the transcript view the first time. The entries can be edited as a table under “Speaker directory” in the sidebar.
What can I do without purchasing?
Transcription, summarization, dictation, and export are all available for free. The one-time in-app purchase “Memotori Pro” removes only these limits, and there is no monthly charge.
| Free | Memotori Pro | |
|---|---|---|
| Audio length | 10 minutes per item | Unlimited |
| Term dictionary | 3 entries | Unlimited |
| Speaker directory | 3 people | Unlimited |
| Summary formats | 2 built-in only | Create, edit, and delete your own |
I bought Pro but the limits are still there
Open Settings (⌘,) > Memotori Pro and choose “Restore Purchases”. Also check that the Apple Account signed in to the App Store is the one used for the purchase.
I would like a refund
Refunds for in-app purchases are handled by Apple at reportaproblem.apple.com. The developer cannot issue them.
Where is my data stored, and how do I delete it?
Everything is stored on your Mac. Transcript and summary history, the dictionary, summary formats, the speaker directory, and models live in ~/Library/Application Support/Memotori/; recordings are saved to the folder you designate. Individual items can be deleted from within the app, and deleting the app removes this data too, except for recordings in the folder you designated.
Which export formats are supported?
TXT, SRT, VTT, Markdown, and JSON. SRT and VTT can be used as video subtitles.
Is my audio or transcript sent anywhere?
No. Recording, transcription, and summarization all happen on your Mac. The only network access is the model download and the App Store purchase. See the privacy policy for details.