CHOROS

The Interface is Everywhere

A ground-up, AI-first operating system. Built on a graph architecture with capability-based security.

“You don't learn the OS,
you just prompt it.”

Before

$ traditional CLI
ffmpeg -i file.mp4 -vn audio.mp3
# Memorize commands
# Parse documentation
# Remember flags and syntax

After

You:
“Computer, pull the audio track from this video”
Understanding intent, processing through graph, executing task...

ChorOS shifts computing from learned skill to natural conversation. The interface is prompt-driven. The OS is cognition.

Core Principles

Three foundational ideas that make ChorOS fundamentally different

PRINCIPLE 1

The Interface is Conversation

Forget memorizing commands, paths, or permissions. ChorOS shifts computing from learned skill to natural conversation. You simply talk, text or voice, and the OS interprets, reasons, and acts. Capabilities and permissions ensure computation is completed safely.

No command memorization required

Voice or text input

Natural language understanding

Traditional:
chmod +x script.sh
./script.sh --output=data.txt
ChorOS:
“Computer, run the script and save output to data.txt”
Graph visualization showing interconnected nodes
PRINCIPLE 2

Everything is a Graph

All system resources - files, processes, devices, and even user sessions - are modeled as nodes in a typed graph. Edges represent relationships and carry capability tokens. This unified abstraction lets the LLM reason about system state semantically and safely.

Unified resource model

Semantic reasoning by LLM

Natural queries: “files I edited last week”

PRINCIPLE 3

Capabilities, Not Permissions

Traditional OS permissions are brittle. ChorOS uses 128-bit unforgeable tokens for access control. Every edge in the graph requires a capability token with rights embedded directly.

128-bit unforgeable tokens

Instant global revocation

Complete audit trails

Principle of least authority by design

Capability Token:
a7f3c9e1b2d4f6a8c1e3b5d7f9a1c3e5
Embedded Rights:
READWRITEEXECUTE
Status:
✓ Active • Audited • Revocable

The Orchestration Under the Hood

Multiple services harmonizing under an AI conductor 🎼

Core Services

Graph Service
The central hub for all resources
Capability Service
Security backbone with unforgeable tokens
Content Storage
Handles large objects like recordings
Audit Service
Transparency and compliance logging
Event Bus
Real-time pub/sub synchronization

Intelligence Layer

AI Gateway
Connect to Claude, GPT, and other models
AI Agent Service
Local inference and session management
Speech I/O
Voice ↔ Text transformation

Human Interface

UI Service
Orb UI, waveforms, visualizations
Credential Manager
Secure OAuth and API key management

Request Flow: “Computer, summarize this audio”

1
Speech I/O
Transcribes your request
2
AI Agent
Parses intent
3
Capability Service
Validates access
4
Graph Service
Routes request
5
AI Gateway
Processes via external model
6
Content Storage
Stores result
7
Audit Service
Logs transaction
8
UI Service
Presents to user

Technical Specifications

Built for performance, security, and flexibility

Languages & Stack

System Services:
Rust
High-Level Skills:
Bring Your Own Language
External Clients:
Broad MCP Support

AI Models

Local Inference:
Local LLMs
Voice Processing:
Voice-to-Text
Cloud Integration:
Claude, GPT, etc.

Hardware Support

Runs on Anything
  • • Raspberry Pi 5
  • • Standard x86_64 systems
  • • ARM64 devices
  • • High-end workstations

Architecture

Core:
Typed graph with capability-secured edges
Skills:
Federated across Rust/Swift partitions
Integration:
Linux VM for POSIX tools
Network:
Local Private Mainframe

Private Mainframe Computing

Local networks orchestrate distributed AI workloads. Each device becomes a node in a larger compute fabric, increasing capabilities and speed while maintaining privacy-first edge computing principles.

Why ChorOS

Built for the future of computing

Local Mainframe Computing

Orchestrate local networks of devices into a unified compute fabric. Privacy-first edge computing with distributed AI workloads.

  • Local network orchestration
  • Distributed AI processing
  • Privacy-first architecture
  • Automatic load balancing

Natural Workflows

Stop memorizing commands. Just describe what you want in plain language and let the OS handle the complexity.

  • Voice or text commands
  • Media processing without docs
  • Semantic file organization
  • Context-aware operations

Developer Experience

Write skills, not apps. Integrate with external services via MCP. Graph-native APIs make complex operations simple.

  • Skills over traditional apps
  • MCP integration support
  • Graph-native APIs
  • Linux VM for POSIX tools

Enterprise Security

Capability-based access control provides fine-grained security by design. Complete audit trails for compliance.

  • Unforgeable capability tokens
  • Complete audit trails
  • Instant token revocation
  • Zero-trust by design

See It in Action

Watch ChorOS running on Raspberry Pi 5

Demo video coming soon

1

Boot ChorOS on Raspberry Pi 5

2

Say: “Computer, convert this recording to text”

3

LLM routes through Linux VM, processes with whisper.cpp

The Pipeline

VoiceLLMGraphResponse

Download ChorOS

Join the revolution in personal computing

Beta Release

Detected platform:

Version
0.1.0-beta

⚠️ Beta Software: ChorOS is in active development. Expect bugs, breaking changes, and frequent updates. Not recommended for production use.

→ Installation Instructions
1. Download the image
curl -O https://releases.choros.dev/choros-latest.img
2. Flash to SD card or USB
dd if=choros-latest.img of=/dev/sdX bs=4M status=progress
3. Boot and follow setup
# First boot will guide you through configuration
→ System Requirements
  • CPU: 64-bit ARM or x86_64 processor
  • RAM: 4GB minimum, 8GB recommended
  • Storage: 16GB minimum, 32GB+ recommended
  • Network: Internet connection for cloud AI models

Stay Updated

Get notified when ChorOS is ready for download and receive development updates

Frequently Asked Questions

Everything you need to know about ChorOS