Understand what you see
Ask questions about an email, message, document, webpage, chart, PDF, or application without copying and pasting it somewhere else.
“What are the three most important things in this document?”
A voice-to-action layer for your computer
Say what should happen next.
Press one key. Speak naturally. Onyday understands what you’re looking at and turns it into a reply, update, answer, or completed action.
FROM SCREEN TO SENT
Skip the blank page. Just say it.
Look at the thread, document, or message in front of you and say what should happen next. Onyday reads the context and turns it into a reply, update, or answer—ready for your approval.
- Understands what’s on screenNo copying, pasting, or describing the context yourself.
- Writes in your voiceReplies sound like you, matched to who you’re speaking to.
- You approve everythingNothing is sent without your confirmation.
From screen to sent—without copying, pasting, or switching apps.
Product demonstration: a Mac desktop with a Gmail thread open. A compact voice interface appears near the top of the screen, transcribes a spoken instruction, reads the visible email thread, drafts a reply, asks for confirmation, and sends it.
Turn anything on your screen into action.
Stop explaining context to your AI. Onyday already understands what you’re looking at, so you can focus on what should happen next.
Communicate in your voice
Draft emails, Slack messages, updates, and replies that match how you communicate and who you are speaking to.
“Reply positively, but make it clear we can’t commit to that deadline.”
Complete the next step
Turn what you are viewing into an email, Slack update, follow-up, or another immediate action.
“Send the main changes to the product channel and mention that the launch date hasn’t changed.”
One key. One sentence. Done.
- 1
Press
Use one global hotkey from anywhere on your Mac.
- 2
Speak
Say what you want in natural language. No commands or special syntax to remember.
- 3
Done
Onyday understands the screen, prepares the result, and completes the action after your approval.
From screen to sent.
Email threadReply
“Tell him Friday works and ask whether the analytics section can wait until next week.”
A personalized reply is prepared in the existing Gmail thread.
DocumentSlack update
“Send the important changes to the launch channel.”
A concise update is created with the source attached.
Customer messageFollow-up
“Apologize, explain that the issue is fixed, and ask whether they can talk tomorrow.”
The response is drafted in the correct conversation.
ChartExplanation
“Explain this and send the takeaway to Alvin without the technical details.”
The chart is interpreted and converted into a clear message for the intended recipient.
Spoken promptAnswer overlay
“Ask Claude to explain this error and how to fix it.”
The response opens in a floating layer above your work—get answers without switching screens.
Drawn regionFocused answer
“Circle this table and tell me what changed since last quarter.”
Only the area you marked is analyzed—the answer stays focused on what you pointed at.
It learns how you work.
Onyday adapts to your writing style, vocabulary, relationships, and workflow preferences—so every action feels less like prompting an AI and more like working naturally.
- How you write emails
- How you communicate in Slack
- Your preferred level of detail
- Frequently used names and terminology
- When you want confirmation before an action
- How your tone changes between teammates and customers
Help shape the future of voice-first work.
We’re inviting a small group of early users to test Onyday and tell us where it saves time, where it fails, and what it should do next.
Early access will be released gradually as the product becomes reliable enough for daily work.
Questions, answered.
What is Onyday?
Onyday is a voice-to-action layer for your computer. Press one key, say what you want, and it uses what is visible on your screen to answer, draft, or complete the next action.
Is this a dictation app?
It can help you write, but it goes beyond dictation. It understands the context on your screen and can help turn that context into an answer, email, message, update, or action.
Can it run my AI prompts?
Yes. Say your prompt out loud and Onyday runs it on Claude, then opens the response as a floating layer on top of whatever you're doing—so you get your answer without switching screens while you work.
Can I ask about just one part of my screen?
Yes. Draw a circle around any area—a chart, a paragraph, a code snippet, a design—and Onyday focuses on exactly that region to answer your question, summarize it, or draft a response about it.
Which applications will it work with?
The first beta will focus on common professional workflows across the browser, Gmail, Slack, documents, PDFs, and other macOS applications.
When will the beta launch?
We are inviting users gradually as the product reaches the level of speed and reliability required for daily work.
Will it take actions automatically?
You remain in control. External actions such as sending an email or posting a Slack message should require clear confirmation.
How do referrals work?
After joining, you receive a personal referral link. When someone new signs up through it, the referral is recorded. People who actively share the product will receive personal follow-up as beta access becomes available.
What should happen next? Just say it.
Join the private beta for the voice-to-action layer for your computer.