Skip to content

README.md

Latest commit

 

History

History
executable file
·
238 lines (154 loc) · 6.22 KB

README.md

File metadata and controls

executable file
·
238 lines (154 loc) · 6.22 KB

🧠 SuperBrain AI – Offline, Ethical, Research-Driven AI Platform (V1.2)

A modular AI research platform combining popular cloud models with offline uncensored LLMs — encrypted, portable, and built for privacy-first users.

🚀 What Is SuperBrain AI?

SuperBrain AI is a local-first, privacy-respecting AI platform built for advanced research, investigation, and offline intelligence. You can run:

✅ API-powered agents like GPT-4, Claude, Gemini, Groq, and Venice
✅ Local LLMs like TinyLLaMA, LLaMA2, Mistral, Phi via Ollama
✅ Dark web scrapers, document ingestion, face recognition, and more (Pro only)

Everything runs fully offline — and every answer is logged, learned, and improved over time through an encrypted memory system.

📥 Download SuperBrain AI

✅ Option 1: Download ZIP (No Setup Needed)

Download ZIP

  1. Click the button above ☝️
  2. Unzip the folder on your computer
  3. Open it and follow the instructions in README.md to get started

🧑‍💻 Option 2: Clone Using Git (For Devs)

🧠 Tip: Make sure you have Git installed

git clone https://github.com/KatchDaVizion/SuperBrain_AI.git
cd SuperBrain_AI
bash launch_ai_portable.sh

🔧 Don't Have Git Yet?

Install Git with one of the commands below:

🐧 For Debian/Kali/Ubuntu:

sudo apt update
sudo apt install git -y

🍏 For macOS (Homebrew required):

brew install git

🪟 For Windows:

  1. Download Git from: https://git-scm.com/downloads
  2. Install and open Git Bash

⚡ Install with One Line (Quick Setup)

git clone https://github.com/KatchDaVizion/SuperBrain_AI.git && cd SuperBrain_AI && bash launch_ai_portable.sh

🔍 Why Use Uncensored Local Models?

Online AI tools are often:

  • Filtered or politically limited
  • Rate-limited or API-bound
  • Logged and monitored by third parties

With SuperBrain, you:

  • Own your data and tools
  • Can run uncensored models like TinyLLaMA or Mistral for raw knowledge access
  • Perform deeper research without censorship
  • Are free to build intelligence with full autonomy

SuperBrain is designed for ethical research only. Misuse contradicts its mission.

🔐 Multi-AI Query Assistant (Research Mode)

SuperBrain includes a unique multi-model querying script:

python3 multi_ai_query.py

🧠 Ask one question → Get responses from GPT-4, Claude, Gemini, Groq, and Venice side-by-side.

Benefits:

  • 🔍 Instantly compare model reasoning styles
  • 🧠 Save all results to memory (RAG-enhanced)
  • 🔒 API keys are encrypted and reused automatically
  • 📊 Tabulated output powered by tabulate

✅ Core Features

  • 🔌 Multi-AI Hub – GPT-4, Claude, Gemini, Groq, Venice (API) + Local LLMs (Ollama)
  • 🧠 Memory System + RAG Recall – Past answers stored and reused semantically
  • 🕵️ Dark Web Scraper – Tor-powered .onion access (Pro)
  • 📚 PDF/Document Ingestion – Feed reports or notes into memory (Pro)
  • 🧑‍💻 Facial Recognition Agent – Match images offline (Pro)
  • 🔐 Encrypted Memory Mode – AES passphrase protection (Pro)
  • 🧹 Plugin Loader – Drop in your own Python tools
  • 📀 USB Portable Launcher – Run from a flash drive or SSD
  • ↻ Versioned, Auto-Updating Architecture

📂 LLM Model Downloads & RAM Estimates

You can download any model from ollama.com directly inside SuperBrain:

/download phi3
/model 2  # switch model
Model Name Minimum RAM Required
TinyLLaMA 1.7 GB
Phi-3 3 GB
LLaMA2 7B 6–8 GB
Mistral 10–12 GB
Mixtral 8x7B 24+ GB

⚠️ Always check available memory before switching models. SuperBrain dynamically adjusts based on system RAM.

🔐 Secure Assistant Setup

All assistants (OpenAI, Claude, Gemini, Groq, Venice) use encrypted API key storage:

  • 🔐 Keys saved to ~/.openai_api.enc, etc.

  • 🔁 Reused securely without plaintext

  • 🧹 Delete all keys with:

    python3 clear_keys.py

💻 Installation

🐧 Linux / WSL / Kali / Ubuntu:

chmod +x launch_ai_portable.sh
./launch_ai_portable.sh

🍏 macOS:

chmod +x launch_macOS.sh
./launch_macOS.sh

🧠 How the AI Learns Over Time

Every assistant (local or API) stores results with:

  • ⏱ Timestamp
  • 📍 Source
  • 🔍 Embedded content for future recall (RAG)

The more you use SuperBrain, the smarter it gets.

🛡️ Ethics Statement

SuperBrain provides uncensored tools. That makes your ethical responsibility vital.

This platform:

  • ✅ Promotes research, privacy, open exploration
  • ❌ Prohibits illegal, unethical, or abusive use
  • 🛡 Defends the right to study and investigate safely

Use it wisely. Empower, don't exploit.

👮 Author & License

SuperBrain AI Platform Created by David Louis-Charles (KatchDaVizion) © 2025 — MIT License

Embedded Signature:

__author_id__ = "KatchDaVizion_2025_DLC_SIG"
def check_license():
    return "David Louis-Charles" in __author_id__

Includes models from ageitgey/face_recognition_models (MIT).

📦 Distribution

🙌 Community & Support

  • 🛠 Issues welcome
  • 🤝 Contributions encouraged
  • 💬 Feedback always appreciated

Built for ethical intelligence by David Louis-Charles.

#OfflineAI #UncensoredLLM #CyberSecurity #OSINT #PrivacyTech #SuperBrainAI #RAG #DarkWebResearch #AI4Good #VeniceAI #PortableAI

⚠️ Note: SuperBrain AI is pinned to compatible LangChain versions to avoid known dependency conflicts.