Build a ChatGPT Alternative with Qwen3 + vLLM + Custom UI

Build a ChatGPT Alternative with Qwen3 + vLLM + Custom UI

Introduction: Your Own AI, No Subscriptions

With Qwen3 and open-source tools like vLLM, it’s easy to create a powerful ChatGPT alternative:

  • 100% local or private cloud

  • Full chat history, streaming, markdown

  • Use any Qwen3 model (7B, 14B, or even 480B)

  • Build with Python, JS, or Gradio

This guide shows how to deploy your own chatbot with:

  • Qwen3 (LLM)

  • vLLM (OpenAI-style API)

  • Simple frontend (HTML/JS or Gradio)

1. Set Up Qwen3 with vLLM

Install vLLM:

bash


pip install vllm

Start the API server:

bash


python -m vllm.entrypoints.openai.api_server \
  --model Qwen/Qwen1.5-14B \
  --port 8000 \
  --enable-token-streaming

Now your local API is running at:

bash


http://localhost:8000/v1/chat/completions

2. Create a Basic HTML Chat UI

Save this as index.html:

html


<!DOCTYPE html>
<html>
<head><title>Qwen3 Chat</title></head>
<body>
  <h2>Chat with Qwen3</h2>
  <div id="chat"></div>
  <textarea id="input" rows="4" cols="50"></textarea><br>
  <button onclick="send()">Send</button>

<script>
  async function send() {
    let input = document.getElementById("input").value;
    const res = await fetch("http://localhost:8000/v1/chat/completions", {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        "Authorization": "Bearer your-key"
      },
      body: JSON.stringify({
        model: "Qwen/Qwen1.5-14B",
        messages: [{ role: "user", content: input }]
      })
    });
    const data = await res.json();
    document.getElementById("chat").innerHTML += "<p><b>You:</b> " + input + "</p>";
    document.getElementById("chat").innerHTML += "<p><b>Qwen3:</b> " + data.choices[0].message.content + "</p>";
  }
</script>
</body>
</html>

✅ Open this in your browser, and chat away—no cloud or account needed!

3. Optional: Use Gradio UI for Zero-Code Interface

python


import gradio as gr
import openai

openai.api_base = "http://localhost:8000/v1"
openai.api_key = "qwen-key"

def chat(prompt, history=[]):
    messages = [{"role": "system", "content": "You are a helpful assistant."}]
    for user, bot in history:
        messages.append({"role": "user", "content": user})
        messages.append({"role": "assistant", "content": bot})
    messages.append({"role": "user", "content": prompt})

    response = openai.ChatCompletion.create(
        model="Qwen/Qwen1.5-14B",
        messages=messages
    )
    answer = response['choices'][0]['message']['content']
    history.append((prompt, answer))
    return answer, history

gr.ChatInterface(chat).launch()

Use this if you prefer no HTML/JS. Just run the script and launch in browser.

4. Why This Beats Hosted LLMs

Feature ChatGPT Your Qwen3 Bot
Cost per use 💸 Paid/subscription ✅ Free after setup
Privacy ❌ Cloud logs ✅ 100% private
Custom fine-tuning ❌ Not allowed ✅ Fully supported
Run offline ❌ No ✅ Yes (local)
Custom API limits ❌ None ✅ Full control

5. Bundle as an App (Optional)

You can bundle your chatbot with:

  • Electron.js (turn HTML into desktop app)

  • Tauri (for mobile/desktop hybrid)

  • Docker (for private cloud deployments)

Conclusion: Your Own AI Assistant—Fully Open

With just:

  • Qwen3

  • vLLM

  • Chat UI (HTML/Gradio)

You’ve built a ChatGPT-level assistant—no API keys, no monthly fees, and no limits.

✅ Customize, fine-tune, or scale however you want. Qwen3 gives you full control.

Resources



Qwen3 Coder - Agentic Coding Adventure

Step into a new era of AI-powered development with Qwen3 Coder the world’s most agentic open-source coding model.

Hugging Face GitHub Modelscope Discord