AI with LM Studio (Local Qwen3-VL-8B)

This boilerplate demonstrates how to use LM Studio to run a fully local LLM environment — hosting the Qwen/Qwen3-VL-8B model with OpenAI-compatible APIs, without paying for tokens.

It provides a complete, end-to-end example of a chat interface powered by a locally hosted LLM, ideal for experimentation, prototyping, and offline development.

Overview

The example integrates Next.js App Router, shadcn/ui, and the Vercel AI SDK (AI Elements) to build a modern chat UI connected to a local LM Studio instance.

A custom tool is included to demonstrate how the LLM can trigger server-side actions, such as generating and saving documents.

This boilerplate is designed as a local AI laboratory, enabling you to test ideas with a real LLM — without network dependency, rate limits, or token costs — before moving to a production-grade hosted model.

Features

LM Studio hosting a local Qwen3-VL-8B model
OpenAI-compatible API for frictionless integration
Next.js App Router for modern routing
Vercel AI SDK (AI Elements) for a polished chat interface
shadcn/ui for consistent UI components
Tool invocation example that creates documents from LLM responses
Fully offline, zero-cost LLM experimentation
End-to-end chat flow with local inference

Use Case

Ideal for developers who want to:

Prototype AI features without paying per token
Run real LLMs locally for fast and private experimentation
Test integrations before switching to cloud LLM providers
Build RAGs, agents, or tools using a local model
Develop AI features in isolated environments with full autonomy

With LM Studio, Next.js, Vercel AI SDK, and Qwen3-VL-8B, this boilerplate provides a powerful foundation for building and testing AI applications — all directly on your machine, without external costs or dependencies. 🚀

Boilerplate