AI Voice Assistant · Windows · Offline-First

Chronos YOUR J.A.R.V.I.S
IS HERE

Control your PC with just your voice. No cloud, no surveillance, blazing-fast local AI that sounds exactly like Stark's right-hand system.

⬇ Download EXE Download Installer

Refer to the User Guide For Commands and How to Use

Core Capabilities

What Friday Can Do

Built for power users who want hands-free control over their machine without sacrificing speed, privacy, or the cool factor.

🎙️

Voice App Launching

Say the name of any application installed on your PC and Chronos launches it instantly. No clicking, no searching just speak and it opens.

🔊

Jarvis-Class Voice

Powered by Kokoro Edge TTS with a carefully selected voice profile that closely mirrors J.A.R.V.I.S. from Iron Man deep, calm, precise, and authoritative.

🧠

Vosk Speech Engine

Uses the Vosk offline speech recognition toolkit, a lightweight, fast, and accurate recogniser that runs entirely on your hardware with no internet required after initial setup.

🔒

Full Local Processing

After the first-use model download, every single word you speak is processed on your machine. No microphone data ever leaves your PC ever.

⚡

Near-Instant Response

Because all processing is local, there's no server round-trip latency. Chronos hears you and acts in under a second, faster than you can reach for the mouse.

📡

Online → Offline Transition

On first run, Chronos Generates the voice and recognition models automatically. Once those are cached, it becomes fully air-gapped, no WiFi needed, ever again.

🪟

Windows 10 & 11 Native

Deeply integrated with the Windows environment. Chronos reads your installed apps, system paths, and startup entries to build its command vocabulary automatically.

🗣️

Natural Language Commands

No need to memorise exact syntax. Say "open Spotify", "launch Chrome", "start VS Code", Chronos understands natural phrasing and multiple command forms.

🧩

Zero Subscription

Download once, use forever. There's no monthly fee, no API key required, no usage limits. Chronos is yours completely and permanently.

Process Flow

How It Works

A streamlined pipeline from voice to action, all running on your machine.

You Speak the Command

Chronos runs quietly in the background, listening via your microphone. When you say the wake phrase followed by an app name, it captures your voice input as a local audio buffer, nothing is streamed anywhere.

Vosk Decodes Your Speech

The Vosk offline speech recognition model processes the audio entirely on your CPU. It converts your voice to text in milliseconds using a compact, locally-stored language model trained specifically for command-style speech.

Intent Matching

The transcribed text is matched against Chronos's internal command registry, which maps natural-language phrases to your installed applications. Fuzzy matching ensures you don't need to say the exact app name every time.

App Launches Instantly

Once a match is found, Chronos fires the corresponding Windows process or shell command. The target application opens, no mouse click, no keyboard shortcut, no context switching required.

Chronos Confirms via Voice

Microsoft Edge TTS synthesises a spoken confirmation using the Jarvis-matched voice profile. You hear "Opening Chrome" or "Launching Spotify", crisp, cool, and immediate. The whole cycle completes in under one second.

Live Demo

Voice Commands

Examples of what you can say to Chronos right out of the box.

Chronos · Voice Session

Hey Chronos, open Chrome

Launching Google Chrome, sir.

Start Spotify

Opening Spotify now.

Open VS Code

Visual Studio Code is ready.

Launch File Explorer

File Explorer opened.

Open Discord

Discord is now running.█

Security & Privacy

Your Voice, Your Machine

Privacy is not an optional feature in Chronos, it's the entire architecture.

⚔ Zero Surveillance Architecture

Chronos was designed from day one around a single principle: your voice data is yours alone. There are no analytics, no telemetry, no usage reporting, and absolutely no transmission of audio or transcripts to any server ever.

All speech-to-text runs on your CPU via Vosk
No microphone stream sent to the internet
No account, login, or registration required
Models stored and run 100% locally
Voice synthesis via local Kokoro TTS engine
No background data collection of any kind
Works fully offline after first-use model download
Open command logs only on your own device

How We Compare

Chronos vs The Alternatives

Feature	CHRONOS	Cortana	Alexa (PC)	Google Asst.
Fully Offline	✔ Yes	✘ No	✘ No	✘ No
No Data Leaves PC	✔ Yes	✘ No	✘ No	✘ No
Jarvis-Style Voice	✔ Yes	✘ No	✘ No	✘ No
Sub-1s Response	✔ Yes	~ Varies	~ Varies	~ Varies
No Subscription	✔ Free	✔ Free	~ Limited	~ Limited
No Account Required	✔ Yes	✘ MS Account	✘ Amazon	✘ Google
App Launch Focus	✔ Core Feature	~ Secondary	✘ Limited	✘ Limited

Under The Hood

The Tech Stack

Chronos is built on a carefully selected stack that prioritises speed, privacy, and minimal system footprint.

Python 3.x

Vosk SR Engine

Kokoro TTS

PyAudio

Windows Shell API

Subprocess Module

Fuzzy Matching

Local NLP Pipeline

Edge Voices API

Win32 App Registry

The Vosk speech recognition library provides compact, high-accuracy offline models that operate with no network dependency. Kokoro TTS delivers the voice synthesis using a voice profile carefully tuned to replicate the iconic J.A.R.V.I.S. tone, measured, intelligent, and instantaneous.

System Requirements

What You Need

🖥️

Operating System

Windows 10 (64-bit) or Windows 11

🎤

Microphone

Any USB, 3.5mm, or built-in mic, no special hardware required

💾

Storage

~400MB for app + voice model (downloaded on first run)

🧮

RAM

4GB minimum · 8GB+ recommended for snappiest performance

🌐

Internet

Required once on first launch to download models. Fully offline after.

⚙️

Dependencies

Python 3.x included in installer. No manual setup required.

Chronos YOUR J.A.R.V.I.SIS HERE