OpenWebUI Integration Guide

This guide shows how to integrate Nexa-SDK with OpenWebUI for a web-based chat interface with locally running models.

Overview

OpenWebUI is a modern, feature-rich web interface for interacting with language models. By integrating it with Nexa-SDK, you get:

Before starting, ensure you have:

First, download the model you want to use:

bash

nexa pull Qwen/Qwen3-VL-8B-Instruct-GGUF

Then start the Nexa Serve API server:

bash

nexa serve

The server will start and listen on http://127.0.0.1:18181. You should see output like:

[INFO] Nexa Serve running on http://127.0.0.1:18181
[INFO] OpenAI-compatible API at http://127.0.0.1:18181/v1

Follow the OpenWebUI Installation Guide to set up OpenWebUI on your machine.

Start OpenWebUI, open your browser and navigate to your OpenWebUI instance, typically at:
```
http://localhost:8080
```
Sign up and login:
- Click the sign-up button
- Create an account with your email and password
- Login with your credentials
Navigate to Settings:
- Click the Profile icon in the top right corner
- Select Admin Panel from the dropdown menu
- Click Settings in the top navigation bar
Configure Nexa-SDK Connection:
- In the left sidebar, click Connections
- Scroll to the OpenAI API section
- Click + Create or Add New Connection
Fill in Connection Details:
- API URL: http://127.0.0.1:18181/v1
- API Base: http://localhost:8080
- API Key: any-key (Nexa Serve doesn't require authentication)
Add Model ID:
- In the Model ID field, enter the model you downloaded, for example:
```
Qwen/Qwen3-VL-8B-Instruct-GGUF
```
- Click Save to add the connection

Problem: OpenWebUI cannot reach Nexa Serve

Solutions:

Problem: The model ID is not recognized

Solutions: