NGARi FAQ

NGARi is a sovereign AI platform that runs entirely on your own hardware. No data leaves your network. No third-party API keys. No content filters or usage policies imposed by someone else. You own the model, you own the data, you control the rules.

Those services run on someone else's servers. Your conversations are stored, processed, and reviewed by their teams. NGARi runs locally on your infrastructure — a Jetson device, a Linux server, a Mac, or any cloud VM. No one sees your data except you. There are no refusal filters and no API rate limits from a third party.

NGARi runs on any machine that can host an Ollama instance. Recommended: NVIDIA Jetson Orin Nano (8GB+), Mac with Apple Silicon, or any Linux server with a GPU (8GB+ VRAM for 14B models, 24GB+ for 32B models). The free tier works on most modern hardware. Larger models (X, Y, Z plans) benefit from more GPU memory.

Yes. All inference happens on your hardware. Conversation data is stored in your own PostgreSQL database. NGARi does not phone home. No telemetry, no tracking, no third-party analytics. The only external network calls are for Stripe billing (if you subscribe) and email sign-in magic links — both of which you can self-host.

The free tier includes Qwen 2.5 Coder 14B, Qwen 2.5 7B, GLM-4 9B, and Llama 3.2 1B. The X plan adds Qwen 2.5 32B. All models run locally through Ollama — no API keys required. You can also add your own models to the Ollama instance.

NGARi includes six specialized AI agents: Sage (Strategy), Maya (Revenue), Kai (Security), Imani (Operations), Zara (Marketing), and Ayo (Coordination). Each has focused system prompts for their domain. Switch between them in Power mode, or use NGARi directly in Simple mode.

The free tier includes 50 conversations per month, access to 4 models, and all 6 agent personalities. No credit card required. Anonymous users get 10 conversations before signing in. Upgrade to X ($79/mo), Y ($249/mo), or Z ($799/mo) for larger models, unlimited conversations, and priority processing.

NGARi uses magic link authentication. Enter your email address and we send a sign-in link. No passwords to remember, no credentials to leak. The email is sent from your own infrastructure — no third-party email service.

Sign in and click Manage in the sidebar to access your Stripe billing portal. You can upgrade, downgrade, or cancel at any time. Changes take effect at the end of your billing period. No lock-in, no cancellation fees.

Yes. NGARi is designed for sovereign deployment. The entire stack — Next.js app, PostgreSQL database, Ollama model server, and email server — runs on your infrastructure. The default deployment uses Docker containers that you can inspect, audit, and modify.

Chat functionality continues to work because models run locally on your hardware. Only authentication (magic link emails) and subscription management (Stripe) require internet access. Your existing session and all local conversations remain fully functional offline.