Talk into your phone. KeptMind listens, sorts your thoughts into clear next steps, and nudges you only when it matters.

features

Voice capture: speak tasks in under 12 seconds

KeptMind voice capture is built for the parking-lot moment — you do not need clean sentences. Hold the mic, speak under twelve seconds, and review sorted tasks in Today. The median capture is 11 seconds, the AI extracts the task, and the original audio is deleted within 24 hours.

Why voice beats typing for ADHD brains

Typing on mobile adds friction when working memory is already loaded. By the time you have unlocked the phone, opened a task app, navigated to the inbox, and tapped into a text field, the original thought has often evaporated. Voice externalizes the thought before you forget the context that made the task urgent.

Short clips keep review fast. The AI sorts intent into actionable steps instead of producing a giant transcript you will never reread. Three to five seconds of speech often beats thirty seconds of typing — both in capture speed and in the quality of the extracted task.

For ADHD adults the difference is not preference, it is whether the thought makes it into a system at all. Most "lost" tasks are not lost in the app — they never reached an app because the friction at the moment of impulse was too high.

How capture works end to end

Press and hold on mobile or web, speak naturally, release. Tasks appear with suggested priority and energy fit. Edit with taps if needed — there is no mandatory wizard. The AI parses dates ("tomorrow at 3"), priority cues ("really need to"), and energy hints ("if I have time").

Text dump uses the same sorting pipeline when silence is required, in meetings, libraries, or shared spaces. The output is identical: a structured task with energy level, optional due time, and a category guess.

Audio is processed in our pipeline, transcribed, parsed into a task, then deleted within 24 hours by default. The text version of the capture stays under your account and travels with you on export.

When voice capture works best

In transit — walking, driving, on public transport. The lock-screen widget on iOS and Android starts recording in under two seconds, no unlock required. This is the highest-leverage moment for ADHD capture, and it is what every productivity app advertises but few actually deliver under real friction.

After meetings or hyperfocus sessions, when ten obligations surface at once. Speak them in sequence as one stream — KeptMind splits them into separate tasks. The follow-up review takes seconds, not minutes.

During phone calls or family chaos, when typing is impossible but the thought is fragile. Voice survives those moments; typing does not. This is why voice is the default modality, not an afterthought.

When NOT to use voice

In quiet shared spaces where speaking aloud is rude — the text dump alternative uses the same sorting pipeline, no feature loss.

For very long planning sessions, use brain-dump mode instead. Voice capture is optimised for sub-twelve-second moments; sustained planning belongs in a different flow.

For tasks that contain sensitive identifiers like medical IDs or passwords. Voice transcription is processed off-device — type those manually if your privacy threshold demands it.

Privacy and data

Audio is deleted within 24 hours after transcription. The text task stays under your account until you delete it or export your data. We do not train third-party models on your voice. We do not sell captures to advertisers.

A 30-day retention opt-in exists for users who want voice notes searchable after capture. It is off by default and configurable in Settings → Privacy. Default deletion is the policy because most people do not think about retention until after they should have.

Frequently asked questions

How long can voice notes be?
Designed for under twelve seconds per capture. Longer rambles belong in brain-dump mode, which accepts bigger input. The twelve-second window is the sweet spot where working memory still holds the context that made the thought urgent — longer clips lose the why even when they save the what.
Does voice capture need a paid plan?
No — core voice capture is on the free tier including AI sorting and Today list integration. Plus adds escalating SMS and call nudges for critical tasks; AI+ adds higher-quota voice processing and call escalation.
What languages are supported?
English at full quality on launch with regional accent support across UK, US, Irish, Australian, South Asian, and African English. We chose to ship one language deeply rather than many shallowly.
Can I use voice capture offline?
Captures recorded offline are queued and processed when connection returns. The audio waits on device, sync happens automatically. You will see a small offline indicator while waiting and a summary when sync completes.
How accurate is the parsing?
On clean speech, task extraction is correct on the first try about 88% of the time, asks one disambiguating question another 8% of the time, and misreads about 4%. The parsed task always shows for review before it goes into Today.
Does it work in noisy environments?
Reasonably well. Modern speech models are robust to background noise — cars, cafes, walking on a street are usually fine. Crowded bars and gyms with loud music are harder. If transcription quality drops, we tell you and let you re-record.
See the productGet the appView pricing

Related

ProductBrain dumpFor ADHD adultsDownloads
Voice capture: speak tasks in under 12 seconds · KeptMind