Skip to content

SyntheticAutonomicMind/SAM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

403 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

SAM - Synthetic Autonomic Mind

A native macOS AI assistant that remembers, gets work done, and keeps your data on your Mac.

I built SAM for my wife. She wanted an AI assistant that could work with her documents, research purchases, generate images, and remember what they'd talked about. Nothing fit, so I built it. SAM was made for her - and dedicated to her.

License: GPL v3 Platform Swift

Website | Download | Part of Synthetic Autonomic Mind


What You Can Do With SAM

SAM is a native macOS app for people who aren't developers. Say "Hey SAM" to go hands-free. Upload a PDF and ask questions about it. Research a purchase across Amazon, Walmart, and eBay in one conversation. Generate images by connecting to ALICE. All without touching a command line.

Your data stays on your Mac. Run local models with MLX or llama.cpp and nothing leaves your machine. Switch to cloud providers when you want more capability - you choose.

Documents & Research

  • Upload and ask questions about PDFs, Word docs, Excel files, and text files
  • Research online with live data from Google, Bing, Amazon, Yelp, and TripAdvisor
  • Create Shared Topics to connect related conversations
  • Search your history by meaning, not just keywords

Writing & Communication

  • Draft emails, essays, or reports
  • Improve your writing with suggestions
  • Brainstorm ideas and organize your thoughts
  • Translate text between languages

Creativity & Design

  • Generate images from text descriptions via ALICE
  • Browse and download models from CivitAI and HuggingFace
  • Apply LoRA adapters for specific art styles
  • Train custom models on your own documents and conversations

Organization & Productivity

  • Manage files and folders with voice commands
  • Create project plans and task lists
  • Organize your work in project folders
  • Access SAM from your iPad or phone via SAM-Web

Math & Calculations

  • Calculate mortgages, tips, compound interest, and ROI
  • Convert units (temperature, length, weight, volume, speed)
  • Get computed answers with Python, not AI approximation

Voice Control

  • "Hey SAM" wake word for hands-free use
  • Speech recognition and text-to-speech
  • Works offline with local models

Personalities

Choose how SAM talks to you - friendly, professional, creative, or create your own.


Screenshots

Natural Conversation

User greeting SAM and asking it to describe what it can help with

Flexible AI Provider Selection

Choose from local models (MLX, llama.cpp), or cloud providers (OpenAI, GitHub Copilot, Google Gemini, DeepSeek, MiniMax, OpenRouter)

Travel Research

User asks about restaurants in a town they'll be visiting

Shopping Assistance

User inquires about a product - SAM helps with research

AI Providers

Provider What You Get
OpenAI GPT-4, GPT-4o, GPT-3.5, o1/o3 models
Anthropic Claude models available via GitHub Copilot and OpenRouter
GitHub Copilot GPT-4o, Claude 3.5, o1 (requires subscription)
DeepSeek Cost-effective AI models
Google Gemini Gemini 2.5 Pro/Flash, large context windows
MiniMax MiniMax-M2.7, M2.5 (128K context)
OpenRouter Access 100+ models from multiple providers
Local MLX Run models on Apple Silicon Macs
Local llama.cpp Run models on any Mac (Intel or Apple Silicon)
Custom Use any OpenAI-compatible API

Switch providers mid-conversation. Use local models for privacy, cloud models for capability. Your choice.


Quick Start

Install

Homebrew (recommended):

brew tap SyntheticAutonomicMind/homebrew-SAM
brew install --cask sam

Manual: Download from GitHub Releases, move to Applications, open SAM.

Set Up Your AI Provider

  1. Launch SAM
  2. Open Settings (,)
  3. Go to AI Providers tab
  4. Click Add Provider
  5. Choose a cloud provider (enter your API key) or a local model (download from within SAM)

Start Talking

Press N for a new conversation. Type your message. Say "Hey SAM" for hands-free.

Access From Other Devices

SAM-Web provides chat from your iPad, iPhone, or any browser. Enable API Server in SAM Preferences, then visit the SAM-Web repository for setup instructions.


System Requirements

To use SAM:

  • macOS 14.0 (Sonoma) or later
  • 4GB RAM minimum (8GB+ recommended)
  • 3GB free disk space

For local AI models:

  • 16GB+ RAM recommended
  • 20GB+ free disk space (models can be large)
  • Apple Silicon (M1/M2/M3/M4) recommended for MLX
  • Intel Macs can use llama.cpp models

Privacy & Security

  • Conversations stored locally in ~/Library/Application Support/SAM/
  • Per-conversation memory isolation
  • API keys stored in UserDefaults
  • Zero telemetry, zero tracking
  • When you use cloud providers, only the messages you send go to those providers

Part of the Ecosystem

SAM is part of Synthetic Autonomic Mind - a family of open source AI tools:

  • CLIO - Terminal AI coding assistant (macOS, Linux, Windows)
  • ALICE - Image generation server (SAM's image engine)
  • SAM-Web - Access SAM from any browser

Documentation

Document What You'll Find
User Guide Getting started, daily use, tips
Features Complete feature reference
Providers AI provider setup and configuration
Tools Built-in tools reference
Memory How memory and search work
Security Privacy and security model
Architecture How SAM is built
Performance Optimization and tuning
Installation Detailed install guide
BUILDING.md Build from source
CONTRIBUTING.md How to contribute
Website Online guides and updates

License

GPL-3.0 - See LICENSE for details.

Created by Andrew Wyatt (Fewtarius) · syntheticautonomicmind.org · github.com/SyntheticAutonomicMind/SAM

Built with open source: Vapor · MLX · llama.cpp · Sparkle