Thoth for ADHD and autism: what I learned

During university, I volunteered with organizations supporting people with physical and cognitive disabilities. It is one of those experiences that stays with you.

I did not build Thoth thinking about accessibility. I was thinking about journalists, lawyers, researchers. People with professional confidentiality needs. Then someone left a review on the App Store:

"I'm neurodivergent. Processing is different in my brain so this removes the need to ruminate and overthink OR remember."

That made me stop. The tool I built for professional privacy turned out to matter for a completely different reason to a completely different group of people. I should have seen it sooner.

Why text is not just a convenience for some people

For most users, a transcript is a shortcut. It saves time. You can skim instead of scrubbing through audio.

For people with ADHD, autistic people, and those with auditory processing disorder or hearing impairment, the difference between audio and text is not always about speed. For many, it is about access.

A transcript you can search, read non-linearly, and return to later is a different kind of record. You control when you engage with it. You can re-read an ambiguous sentence without the speaker having already moved on. The information is there when you are ready for it, not only in the moment it was said.

For people who are hard of hearing or deaf, Thoth's live transcript panel shows words on screen as they are spoken. Privately, locally, with no audio going anywhere.

None of this requires Thoth to do anything different from what it already does. The features are the same. What changes is who they matter to, and why.

The privacy argument is sharper here

I have written before about why cloud recorders create legal exposure for professionals. The argument for accessibility users is simpler and more personal.

A therapy session is not something you want on a third-party server. Neither is a psychiatric evaluation, a medical consultation about a diagnosis, or a conversation with a specialist about a disability accommodation. These are among the most sensitive recordings anyone would ever make.

Cloud transcription services receive that audio. It sits on their infrastructure, processed under their retention policy, subject to their security posture and legal obligations. Some of them have humans review audio samples for quality assurance. Some have been breached. All of them know something about you that you may not have intended to share with a company.

Thoth does not. Recording and transcription are always on-device. Thoth has no servers. If you use AI summaries, a local model option keeps everything on your Mac. If you prefer a specific AI via your own API key, only the transcript text goes to that provider. The audio stays on your device, and Thoth is not in the chain either way.

For most professional use cases, this is a compliance argument. For accessibility users recording medical and mental health appointments, it is a dignity argument.

The iPhone angle

When I explored what a Thoth iOS app would look like, the main limitation I identified was system audio. iPhone cannot capture the audio from a Zoom or Teams call the way Mac can. So an iPhone version would only record your microphone, which makes it a different tool for a different use case.

The more I think about accessibility users, the more that limitation disappears.

The use cases that matter most for this group are in-person: doctor appointments, therapy sessions, lectures, job interviews, conversations you need to be able to go back and read. These are all microphone recordings. The system audio limitation does not apply.

An iPhone version of Thoth would actually be a better fit for these use cases than the Mac version in some ways. Your phone is with you at the appointment. Your Mac is not. Recording on the device you already have in your pocket, getting a live transcript on screen you can follow in real time, and having the full text waiting when you get home, that is a workflow that makes sense.

The iCloud bridge idea I sketched out in the iOS post fits naturally here too. Record on iPhone in the clinic or lecture hall, open Thoth on Mac later to run the higher-quality batch transcription and speaker detection, and have the result sync back. The audio stays within your own Apple account the whole time. Thoth's servers are not in the chain because Thoth has no servers.

I am not announcing an iPhone app. The Mac version is still the focus. But this is now a clearer reason to build one when the time comes.

What I added to the website

I added an accessibility page that names these use cases directly: ADHD, autism, hearing impairment, dyslexia, auditory processing disorder, non-native speakers. The copy I had before was too vague. People searching for a transcription app that works for their specific situation should be able to find it.

What comes next

I want to keep making Thoth more useful for people with accessibility needs, and I am genuinely not sure yet what that looks like in practice. Better keyboard navigation. Larger text options in the transcript view. Faster export flows for people who need to get to their reading app quickly. I am open to things I have not thought of.

One thing I feel strongly about: I do not want accessibility use to be gated behind a Pro subscription. If someone is using Thoth to get through a medical appointment or process a lecture, that is exactly the kind of use I want to support. My intention is to make sure the features that matter most for accessibility are available on the free tier. I will work toward that as the app develops.

If you use Thoth for accessibility reasons and have a suggestion or a frustration, I want to hear it. The support page has my contact details. I read and reply to everything personally.

---

Thoth is a private meeting recorder for Mac. All audio stays on your device. Built by one person, no funding, no team. If you find it useful, upgrading to Pro is the best way to support development.