Work at the speed
of thought
For people who think faster than their tools. Kensi connects your voice to everything you’re working on. You can now dictate, ask, rewrite, and capture meetings with AI that understands your screen, your tools, and your workflow.
macOS · iPhone · Free to start
“
The problem
Thinking. Typing. Switching. Rewriting.
Meeting. Returning. …Wait, What Was I Doing?
Your ideas are faster than your fingers
You think in full sentences. Your keyboard turns them into keystrokes. By the time you’ve typed it out, you’ve lost half of what you meant to say.
Voice tools don’t speak your language
Say "kubectl" and they hear "cube cuddle." Say "PostgreSQL" and they give you "post press." Current voice tools weren’t built for how you actually talk.
Decisions live in five different apps
The answer was in Slack. The spec is in Notion. The meeting happened on Zoom. You spend more time looking for context than using it.
Meetings disappear the moment they end
Action items, decisions, the thing someone said in minute 23 - gone. Meetings are where decisions go to hide. Unless you spent the whole time typing notes instead of talking.
If you threw away your keyboard and mouse, Kensi is how you’d still get everything done.
Dictate into any app. Rewrite any text by voice. Ask questions about what’s on your screen. Record every meeting. Kensi connects your voice to your work - and understands both.
What you can do
For people who think
faster than their tools
Speak. Kensi writes.
Hold a key and talk. Kensi turns your speech into clean text - grammar fixed, filler words gone, technical terms right. Then pastes it wherever your cursor is.
Slack, Gmail, VS Code, Notion, anywhere. Say "um" and "uh" as much as you need. Kensi handles it.
Select it. Say what you want. Done.
Highlight any text, hold Fn, and tell Kensi what to do. "Make this more professional." "Translate to Spanish." "Simplify." The text is replaced right where it is. You never leave the app.
Ask your Kensi. It already knows.
Ask a question by voice. Kensi already has the context - your screen, your files, your Slack, your last meeting. No copy-pasting. No "let me explain what I’m looking at."
It knows what you’re working on. It knows what was said ten minutes ago. Just ask.
@david - agreed, eng can finish the component lib by then
@sarah - works for me. updating the timeline now
Meeting notes without the meeting bot.
Kensi listens in the background - your mic, system audio, or both at once. It knows who said what. And when the meeting ends, you get a summary, action items, and key decisions.
Then ask Kensi about what was just discussed. "What were the action items?" It already heard everything.
Decision: Prioritize auth migration (OAuth only). SAML deferred to next sprint.
Action: David to complete token migration by Thursday.
Blocker cleared: Three teams unblocked once OAuth ships.
Say the word. Get the template.
Say "standup update" and get a full formatted template instantly. Say "bug report" and the structure appears. No AI processing needed - it’s instant.
The difference
Other voice tools hear words.
Kensi understands your work.
Say "what’s wrong with this?" while looking at code. Kensi already knows what’s on your screen. It understands the context and answers without you pasting or explaining anything.
max: 20 is sufficient for your concurrency — if you’re exceeding 20 simultaneous connections, new requests will queue until a connection frees up or the timeout hits.Screen context
Kensi understands what’s on your screen. Which app, which file, which function - it already knows.
File context
Aware of your project’s file tree. Say "open utils dot ts" and Kensi finds the right file.
Visual context
In Ask mode, Kensi understands your current window. Say "what’s this?" and it knows what you mean.
What’s different
Built different.
Not just better.
Most voice tools stop at the transcript. Kensi starts there - then adds context, vocabulary, and understanding that makes the output worth using.
Knows your role
10 personas shaped to how different roles speak. A developer says "post gress" and gets PostgreSQL. A sales rep says "two fifty k" and gets $250K. Same app. Different brain.
Gets terms right the first time
850+ terms built into how Kensi hears - not corrected after the fact. Even the $2/mo tier gets technical terms right on the first pass.
Connected to your tools
Slack, GitHub, Linear, Notion, Google Workspace. Kensi brings in context from the tools you already use - detected from what you’re working on.
Private when you need it
Everything can run on your Mac. Speech recognition, AI correction - all local, all on Apple Silicon. Nothing leaves your machine. Not even metadata.
Works everywhere you type
Menu bar on Mac. iPhone app. System keyboard. Kensi pastes into any text field on any app. Nothing to configure.
Voice snippets
78 voice-triggered text expansions. Say "standup update" and get a full formatted template instantly. No AI needed. No delay.
Built for your role
Kensi speaks your
language
Pick your role during setup. Kensi shapes its vocabulary and formatting to match the way you talk about your work.
Create unlimited Personas with a custom emoji, name, and description. Each has its own fully editable prompts for Dictate, Rewrite, and Ask — independently. Medical, legal, academic, creative, or anything specific to your team.
How it works
Simple to use.
Three steps.
Hold a key and speak
On Mac, hold a key - speak, and Kensi writes into whatever app you’re using. On iPhone, tap the mic. Talk naturally. Filler words, self-corrections, mid-thought changes - Kensi handles all of it.
Kensi understands the context
Your screen. Your files. Your connected tools. Your role’s vocabulary. Kensi uses all of it to understand what you mean and format it right.
Clean text appears
Grammar corrected. Filler words removed. Technical terms spelled right. Pasted exactly where your cursor is - or answered with context from your tools and meetings.
Pricing
Start free.
Pay for more.
No trials. Pick the plan that matches how you work.
- ✓Clean, corrected dictation
- ✓850+ technical terms, 32 packs
- ✓Shaped to your role
- ✓Voice snippets
- ✓Mac + iPhone
- ✓Unlimited words
- —No AI rewrite or ask
- —No on-device processing
- ✓Everything in Voice
- ✓AI correction powered by cloud
- ✓Dictate, Rewrite, and Ask modes
- ✓Personas shaped to your role
- ✓Aware of your screen and files
- ✓Ask mode with screen context
- ✓Live Notes for live meetings
- ✓Audio history
- ✓Unlimited words
- ✓Everything in Voice
- ✓On-device AI correction (Mac)
- ✓Rewrite & Ask modes
- ✓Screen context + workspace awareness
- ✓Live Notes for any meeting (online & in-person)
- ✓Works without internet
- ✓Nothing leaves your machine
- ✓Unlimited words
- ✓Cloud + Privacy combined
- ✓Switch cloud/on-device per mode
- ✓All intelligence modes
- ✓Connected to Slack
- ✓GitHub, Linear, Notion - coming soon
- ✓Google Workspace - coming soon
- ✓Priority support
- ✓Unlimited words
From real users
Once you try it, typing
feels slow
“I dictate every Slack message now. It gets technical terms right that Apple Dictation butchers every single time.”
“The meeting transcription alone is worth it. I stopped taking notes during standups and I actually participate now.”
“I asked Kensi what was decided in Slack while looking at the thread and it just… answered. With action points. From the actual conversation.”
Slack live today. Others shipping soon.
Frequently asked
Apple Dictation gives you a raw transcript. Kensi gives you corrected, formatted text - with vocabulary that knows your field, filler words removed, and self-corrections handled. It also does a lot more: rewrite text by voice, ask questions with full context from your screen and tools, and capture meetings with summaries. Different thing entirely.
Yes. The Privacy plan runs everything on your Mac - speech recognition and AI correction both happen locally on Apple Silicon. Nothing leaves your machine. The Voice plan also works without cloud AI because the vocabulary is built into how Kensi hears. Only the Everything plan’s cloud features need internet.
All of them. On Mac, Kensi lives in your menu bar and pastes into whatever app has focus - Slack, Gmail, VS Code, Cursor, Notion, Google Docs, Terminal, anything with a text field. On iPhone, the Kensi keyboard works everywhere on your device.
Kensi is aware of the text and content on your screen - which app you’re in, what file you’re editing, what’s visible. In Ask mode, it also understands what’s in your active window visually. So you can say "what does this error mean?" or "summarize what’s on screen" and Kensi has the context without you explaining anything. Works with VS Code, Cursor, Xcode, Zed, Terminal, iTerm, Warp, and most other apps.
No. Live Notes captures audio from your mic and system audio directly - no bot joins the Zoom or Meet call. Nothing shows up for other participants. It’s just Kensi running quietly on your Mac.
Yes. Three paths: cloud models on Kensi’s servers (fast and full-featured), a model that runs locally on your Mac’s Apple Silicon (fully private - nothing leaves your machine), or Apple Intelligence as a free fallback on macOS 26+ and iOS 26+. You pick the balance between speed, capability, and privacy.
Your keyboard just became
optional
Stop typing. Start talking. Kensi is ready when you are.
Free to start on macOS and iPhone