Getting Started with Nosana
This guide will help you choose the best way to interact with Nosana based on your needs.
What Do You Want to Do?
Deploy AI Workloads
If you want to run AI models, inference jobs, or other GPU workloads on the Nosana Network, choose one of these methods:
Easiest - No Code Required
The web-based dashboard is perfect for getting started quickly without writing any code.
- Best for: Visual management, quick deployments, learning the platform
- Get started: Visit deploy.nosana.com
- Features:
- Create and manage deployments visually
- Monitor job status and logs
- Manage credits and billing
Host GPUs
If you want to earn $NOS by providing GPU resources to the network:
- Get started: Host GPUs Guide
- Requirements:
- NVIDIA GPU (CUDA compatible)
- 12GB+ RAM
- 256GB+ NVMe SSD
- Ubuntu 20.04+ (Linux recommended)
- Earnings: Earn $NOS tokens for running jobs on your hardware
Understanding the Basics
Before you start, it's helpful to understand a few key concepts:
- Jobs - Individual compute workloads that run on the network
- Deployments - Orchestration layer that manages job lifecycles
- GPU Markets - Pools of GPU resources where jobs are scheduled
- Credits - Prepaid credits used to pay for compute resources
- Hosts - GPU machines that run your jobs on the network
Learn more in the Key Concepts guide.
Authentication Methods
Depending on your chosen method, you'll need different authentication:
- Dashboard: Google login or Solana wallet
- API/SDK with API Key: Get your API key from deploy.nosana.com → Account page
- API/SDK with Wallet: Use wallet-based authentication (see Wallet Authentication)
- CLI: API key or wallet
- Blockchain Programs: Solana wallet
Next Steps
Once you've chosen your path:
- Read the relevant documentation for your chosen method
- Follow the quick start guide for your use case
- Explore examples in the documentation
- Join the community for support and updates
Need Help?
Join our Discord server to get help from the community and the Nosana team.
Ready to get started? Choose your path above and dive into the documentation!