5.4 KiB
GT AI OS Community Edition
A self-hosted AI platform with data privacy for individuals and organizations. Build and deploy custom generative AI agents and bring-your-own local or external API inference via NVIDIA NIM, Ollama, Groq, vLLM, SGLang and more.
GT AI OS is ideal for working with documents and files that need data privacy. Ensure that you are using local or external inference with sero data retention if you want your data to remain private.
Supported Platforms
| Platform | Host Architecture | Status |
|---|---|---|
| Ubuntu Linux 24.04 | x86_64 | |
| NVIDIA DGX OS 7 (Optimized for Grace Blackwell Architecture) | ARM64 | |
| macOS (Apple Silicon M1+) | ARM64 |
Ubuntu VM's running on Proxmox with raw all functions GPU passthrough works. Windows is currently not supported. macOS x86_64 support is being considered although it will be quite slow.
Note that the install scripts are unique for each OS and hardware architecture.
Embedding model GPU acceleration: Only NVIDIA GPU's are supported for embeddings. Ensure that your NVIDIA GPU hardware is installed prior to starting the GT AI OS installation. NVIDIA drivers and dependencies and tools will be installed during the pre requisites part of the runbook.
If you do not have an NVIDIA GPU in your target install host, then the CPU will be used for running the embedding model. CPU vs GPU accelerated embedding will exhibit slow file uploads when adding files to datasets
At v2.0.34, once you install GT AI OS, you cannot install GPU hardware and switch from CPU to GPU for embeddings. We are looking to fix this in a future release. Embedding model is installed by default.
Features
- AI Agent Builder - Create custom AI agents with your own system prompts, categorization, role base access and guardrails
- Local Model Support - Run local AI models with Ollama (completely offline)
- Document Processing - Upload documents into datasets and create agents to interact with them
- Create Teams - For setting up a workgroup that has Team based access to agents and dataasets
- Observability - See metrics dashboards including agents, models and dataset usage, chat logs and more
Documentation
| Topic | Description |
|---|---|
| Installation | Detailed setup instructions |
| Updating | Keep GT AI OS up to date |
| NVIDIA NIM Setup | Enterprise GPU-accelerated inference |
| Ollama Setup | Set up local AI models |
| Groq Cloud Setup | Ultra-fast cloud inference |
| Cloudflare Tunnel | Access GT AI OS from anywhere |
| Troubleshooting | Common issues and solutions |
Community vs Enterprise
| Feature | Community (Free) | Enterprise (Paid) |
|---|---|---|
| Users | Up to 50 users | User licenses per seat |
| Support | GitHub Issues | Dedicated human support |
| Billing & Reports | Not included | Full financial tracking |
| Pro Agents | Not included | Pre-built professional agents |
| AI Inference | BYO/DIY | Fully Managed |
| Setup | DIY | Fully Managed |
| Uptime Guarantee | Self | 99.99% uptime SLA |
Want Enterprise? Contact GT Edge AI
Architecture
┌────────────────────────────────────────────────────────────────┐
│ GT AI OS │
├──────────────────┬──────────────────────┬──────────────────────┤
│ Control Panel │ Tenant App │ Resource Cluster │
│ (Admin UI) │ (User UI) │(AI Inference Routing)│
├──────────────────┴──────────────────────┴──────────────────────┤
│ Postgres DB │
│ Control DB │ Tenant DB │
└────────────────────────────────────────────────────────────────┘
Contributing
Found a bug? Have an idea? Open an issue: https://github.com/GT-Edge-AI-Internal/gt-ai-os-community/issues
See CONTRIBUTING.md for details.
Security
Found a security issue? Report via our contact form
See SECURITY.md for our security policy.
License
Apache License 2.0 - See LICENSE
GT AI OS Community Edition | Made by GT Edge AI