Skip to main content

Quick Start Guide

This guide helps you complete the essential setup flow from accessing the platform to running and serving your first AI model. Follow the steps below to get started quickly. Based on your onboarding flow and uploaded guides.

Step 1: Sign In to the Platform​

Access the Loc.ai Control Platform from your browser, enter your credentials, and open the dashboard where you can manage devices, models, and inference activity.

Step 2: Register Your First Device​

Register your edge device to connect it with the platform. Once the device is online, it becomes available for model deployment and local inference.

Step 3: Deploy Your First Model​

Open the Models section and deploy either a shared pre-built model or your own uploaded custom model to the registered device. Once deployment is complete, the model is ready to run.

Step 4: Run First Inference​

Go to the Inference section, start your deployed model, and send input data such as text, images, or sensor values. Review outputs, confidence scores, and execution details in real time.

Step 5: View and Export Results​

Use the Inference Results page to review logs and predictions. Apply filters, choose columns, and export results as a CSV file for reporting or analysis.

Step 6: Serving a Model​

Enable serving mode to keep your model continuously available for live requests. Loc.ai provides a local OpenAI-compatible endpoint that can be used with tools like VS Code, Obsidian, Open WebUI, and more.

You are now ready to build and run AI workloads locally with Loc.ai.

If you have already installed LocAI Link, you do not need to register your device again. If you closed the Loc.AI Link terminal or exited the application, your device will go offline and become unavailable for inference. To make your device active again, navigate to the installation directory where you installed the Locai:link in your machine and run the startup command again to reconnect your device to the platform.