Skip to main content
Live on your desktop right now →

Just talk to your computer. It gets things done.

Say “Pulse”, describe what you want in plain language, and it plans the steps, does the work, and narrates every move out loud — so you're never left guessing what your computer just did.

Built accessibility-first, for anyone who'd rather speak than click.

View on GitHub Runs 100% locally No cloud round-trip Fully open source

Technology should adapt to people—not the other way around.

Modern computers expect users to learn shortcuts, menus, commands, and complex interfaces before getting real work done.

For many users—especially people with visual impairments or limited mobility—those barriers can make everyday tasks frustrating.

Pulse changes that.

Instead of learning how to use software, users simply describe what they want to accomplish.

How Pulse Works

1

Wake

Simply say: "Pulse." Pulse wakes up and responds naturally. "I'm listening."

2

Understand

Pulse understands your intent rather than requiring exact commands. "Open my electricity bill." "Launch Chrome." "Find my resume."

3

Plan

Pulse breaks your request into manageable actions and decides the safest path to complete the task.

4

Execute

Pulse performs each step while continuously informing you about its progress.

5

Confirm

Every sensitive action requires your approval before execution.

This is what talking to your computer actually sounds like.

No commands to memorize, no menus to hunt through. Four real exchanges — this is the entire interaction model.

Natural language

You

"Pulse, open my email folder."

Pulse

Opening your email folder now.

Multi-step tasks

You

"Pulse, open Notepad and write my grocery list."

Pulse

Notepad's open — go ahead, I'm typing as you talk.

Screen-aware

You

"Pulse, what does this button do?"

Pulse

"That's Send Invoice. Want me to click it?"

You stay in control

You

"Pulse, delete this file."

Pulse

"This permanently deletes report_draft.docx — should I continue?"

Core Principles

Accessibility First

Every interaction is designed to reduce uncertainty by keeping users informed about what Pulse is doing.

Voice First

Voice is the primary interface. No complicated keyboard shortcuts are required.

Local & Private

Pulse is designed to run locally using open AI models whenever possible. Your personal information stays on your computer.

Human Control

Pulse never silently performs important actions. You always remain in control.

Beyond screen readers

Screen readers describe your screen. Pulse operates it.

NVDA and JAWS have been the backbone of computer accessibility for over 20 years — read the screen out loud, then leave you to operate it yourself: tab through elements, memorize dozens of key combinations, mentally map an app before you can act on it. Pulse was built to replace that entire model, not sit on top of it. And because it's just a conversation, not a tool you have to learn, anyone can use it to run their computer by voice.

Screen readers narrate. Pulse acts.

NVDA and JAWS read out labels, buttons and menus so you know what's there. Pulse listens for what you want to accomplish and does it — you describe the goal, not the sequence of clicks to reach it.

No new interface to learn

JAWS alone has dozens of commands, often different between desktop and laptop layouts. Pulse only asks you to talk, in the words you'd already use with another person.

No license fees, ever

A JAWS professional license can run into four figures, and even the cheaper annual home plan is a recurring cost. Pulse is free and fully open source — no subscription, no upsell, ever.

Built to replace, not bolt on

Pulse isn't another add-on running next to your screen reader. It's built to be the whole interface — understand what's on your screen, understand what you want, and do it, through one voice conversation.

What Pulse Can Do

Open applications
Open files and folders
Understand natural language requests
Know what's on your screen to resolve requests correctly
Search for documents
Execute multi-step computer tasks
Speak progress naturally
Continue conversations without repeating the wake word
Confirm important actions before execution

🦸 Superhero Mode

Designed for users who want continuous guidance.

Most software becomes silent once an action starts.

Pulse doesn't.

Superhero Mode provides continuous spoken feedback so users always understand what is happening.

It can also announce typing, navigation, loading states and important system events, giving users continuous awareness of their computer instead of moments of silence.

Silence is disorienting when you can't see the screen — a few quiet seconds after clicking something can feel like it broke. Superhero Mode exists so that never happens.

Never wonder if it froze — hear every step, live
Learn what things are as you go, not just what changed
One switch does it all — no five settings to hunt for

Example Interaction

User

Pulse, enable Superhero Mode.

Pulse

Superhero Mode activated. I'll guide you through everything I do.

User

Open my Downloads folder.

Pulse

Opening Downloads.

I'm reading the folder contents.

I found 24 items including Documents, Images and Videos.

What would you like to do next?

Built Completely with Local AI

Pulse demonstrates what modern local AI can accomplish without relying on cloud infrastructure for core functionality.

Using open models and local inference, the project focuses on delivering practical AI experiences that are private, transparent, and accessible.

Also from the same builder

Local AI Showcase: Animatron

Animatron demonstrates the same local-first philosophy by generating AI-powered animations entirely on consumer hardware. Everything shown here is generated locally without requiring cloud GPU services.

Examples generated purely on local hardware:

Character generation
Scene generation
Image-to-video animation
Local AI rendering pipeline

This showcases the broader vision of practical, accessible AI running directly on personal computers.

Roadmap

We're building Pulse incrementally with carefully tested features.

Open source means you're not waiting on a vendor's release notes — you can watch this happen, or help build it.

Current Foundation

  • Voice interaction
  • Wake word
  • Local AI planning
  • Basic computer control
  • Accessibility feedback
  • Better file understanding
  • Longer multi-step tasks
  • Smarter planning
  • Guided accessibility workflows
  • One-click Windows installer
  • Instant, event-driven screen awareness

Future

  • Browser automation
  • Developer workflows
  • Cross-platform support
  • Plugin & Skill ecosystem
  • Rich computer understanding

Frequently Asked Questions

Everything people ask before they try Pulse.

Yes — that's the whole point. NVDA and JAWS are built around one idea: read the screen out loud and let you navigate it manually. Pulse replaces that model entirely — describe what you want, and it plans and does the task itself. It's accessibility-first by design, but that same simplicity means anyone can use it to control their computer by voice, not only people who currently rely on a screen reader.

Built with Open Technologies.

Pulse is powered by open technologies including local language models, speech recognition, text-to-speech and wake-word detection.

The project exists to demonstrate that capable AI assistants can be built while respecting user privacy and keeping users in control.

Explore Pulse

See how Pulse works, explore the source code, and follow its development as we build a more natural way to interact with computers.

Help Shape Where Pulse Goes Next.

Feedback, bug reports, and contributors are all welcome. Send a message below and it goes straight to the maintainer.

Pulse

A local AI computer companion focused on accessibility, privacy and natural interaction. Built using local AI and open technologies.

© 2026 Pulse Project