Voice input on your phone

Create tasks by voice – just say what needs doing

Something comes to mind on the go, your hands are full, typing takes too long? Tap the plus in Todowa on your phone and say it: “Create a to-do for tomorrow at 8:30: send the invoice to the agency.” Todowa listens until you tap “Done”, transcribes with AI speech recognition and turns it into a task with date, time and project – or an event in your calendar.

Start for free – add Pro anytime

Video shows the German interface; English videos are coming.

How to add tasks by voice

The plus sits in the bottom bar of the phone view. Tap it briefly and Todowa listens – until you tap “Done”. Take your time: pauses to think don’t end the recording. Then the AI creates the entries and shows you what it made.

Tap the plus, speak, then tap “Done” – up to five minutes at a time
The list shows what was created, with date and time; tap a task to open it
“Undo” takes it all back, “Something else” starts listening again right away
Prefer typing? “Type” opens a text field – the AI understands the sentence just the same

Video shows the German interface; English videos are coming.

AI speech recognition that knows your names and projects

Ordinary speech recognition doesn’t know “Website relaunch”, “Ms Kaya” or your “Taxes” tag and writes something that only sounds similar. So Todowa records itself and has the recording transcribed by AI speech recognition (gpt-4o-transcribe by OpenAI). As hints it gets the names from your workspace: people, projects, tags and the titles of your latest tasks.

Then the AI that creates the tasks reads the text again with the same knowledge and fixes mishearings: if a word sounds like one of your projects, that’s probably what you meant. Slips and self-corrections (“at eight, no, at nine”) are taken the way you meant them. You see what was understood below the answer.

Where AI speech recognition isn’t available, your browser’s speech recognition takes over. Todowa still listens until “Done”, and the AI still fixes mishearings using your names and projects.

Ask your to-do system

What the AI makes of what you say

No commands to learn. Say it the way you’d tell a colleague – the AI picks out what goes where:

“tomorrow”, “on Friday”, “half past eight” become date and time
“one hour” becomes the duration
Name a project you already have and the task goes there
“important” or “urgent” sets high priority
Further details and a mentioned place go into the description
Several things in one sentence: up to five entries at once

The inbox AI and Todowa’s vision

Add events to Google or Outlook Calendar by voice

If you explicitly say “appointment”, “meeting” or “event” – say, “Put a meeting with Anna in my calendar tomorrow at 10” – Todowa creates an event in your connected calendar: in Google Calendar if it’s connected, otherwise in Outlook. Without a duration it lasts one hour.

If no calendar is connected, it becomes a task with a time, and Todowa tells you so. Anything with “remind me” or “I need to” stays a task, even with a time. Tasks without a date land in your inbox.

Outlook and Google Calendar Tasks in Google Calendar

Long press: task, event or voice input

Hold the plus down and a small menu opens: “Task” opens the usual quick add, the calendar entry opens quick add for an event in your connected calendar, and “Voice input” starts listening.

If your browser can neither record nor recognise speech, a short tap on the plus opens “New task” as before – and voice input still works with typed text.

Who’s listening? Speech recognition and privacy

With AI speech recognition the recording goes to OpenAI together with names and terms from your workspace (people, projects, tags, titles of your latest tasks), and OpenAI turns it into text. Todowa doesn’t store the recording; according to OpenAI, data sent through its API isn’t used for training.

Without AI speech recognition your browser or operating system converts speech to text – with Chrome and Android usually Google, with Safari and iOS Apple – and Todowa only receives the text.

Todowa sends the text with today’s date, your time zone and the same terms to the AI (Anthropic or your own AI endpoint) so it can create tasks or events. All of this only happens when you start voice input. Details are in the privacy policy.

More than your keyboard’s dictation

Your phone’s dictation only writes down what you say. In Todowa it becomes a scheduled task straight away: with a due date, in the right project and – if you’ve turned on reminders – with a notification on your phone and watch when it’s due. On your computer you’ll find it in the same list.

Reminders To-do list app for Android

Why we’re building Todowa

Your tasks belong to you – not to your tools.

Tasks are spread across many tools, and none of them sees the whole picture. We’re building Todowa as an AI-first to-do system that fits into the ones you already use: tasks show up in your Notion database, in Asana, in Todoist and in Google Calendar, and AI matches up what belongs together.

Our goal for Todowa

What voice input can’t do yet

So you know what you’re getting:

Creating by voice only on your phone: the plus with voice input sits in the bottom bar of the phone view. On the computer you add tasks with the keyboard (Q key); there you can only speak questions to “Ask your to-do system”.
Todowa doesn’t talk back – the answer appears as text on screen.
Listening only after a tap: no “Hey Todowa” and no connection to Siri, Google Assistant or Alexa.
Whether you can speak depends on the browser: it has to allow the microphone. Without AI speech recognition it also needs its own speech recognition (Firefox, for example, has none) – otherwise you type the sentence.
German or English is recognised, depending on the app language. A recording lasts at most five minutes. Voice input doesn’t work offline.
Create only: changing, moving or completing existing tasks by voice isn’t possible. Voice input doesn’t set tags, assignments to others, repeats or new projects.
Events go to the default calendar (Google, otherwise Outlook); you can’t pick another calendar by voice.
Despite your terms, the AI may mishear or misunderstand. That’s why Todowa shows what was created afterwards, and “Undo” takes it back.

Frequently asked questions

How do I create tasks by voice?

Open Todowa on your phone, tap the plus at the bottom centre and say what needs doing – for example “Send the quote tomorrow at 9, project Marketing” – and tap “Done”. The AI creates the task with date, time and project and shows it to you.

Can I add events to my calendar by voice?

Yes, if your Google Calendar or Outlook is connected (Settings → Calendar). Say “appointment”, “meeting” or “event” explicitly. Without a connected calendar it becomes a task with a time.

Does Todowa store my voice?

No. With AI speech recognition the recording goes to OpenAI for transcription and isn’t stored by Todowa. Without it, your browser or operating system (usually Google or Apple) does the recognition and Todowa only receives the text.

Does it work on iPhone?

Yes. Allow Todowa to use the microphone in Safari; transcription uses AI speech recognition, otherwise Apple’s. If the microphone is blocked, type the sentence into the text field – the AI creates the entries just the same.

Is voice input available on the computer?

Not for creating tasks – so far only in the phone view. On the computer you add tasks with “New task” or the Q key. You can speak questions about your tasks there, though – via “Ask your to-do system” in the header bar.

Does voice input cost anything?

It’s included in every plan and, like the other AI features, counts towards the monthly AI allowance; the free plan includes an AI allowance. With your own AI endpoint creating the entries doesn’t count towards the allowance, but AI speech recognition does.

Bring your tasks together.

Create an account, connect your tools, get going – free, add Pro anytime.