How to stream LlamaIndex response in web app

pip install llama-index openai fastapi uvicorn

import os
import asyncio
from fastapi import FastAPI, Request
from fastapi.responses import StreamingResponse
from llama_index import GPTSimpleVectorIndex, SimpleDirectoryReader, ServiceContext
from openai import OpenAI

app = FastAPI()

# Initialize OpenAI client with streaming enabled
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

# Load documents and build index (example with local docs)
docs = SimpleDirectoryReader("./docs").load_data()
service_context = ServiceContext.from_defaults(
    chunk_size_limit=512,
    llm=client
)
index = GPTSimpleVectorIndex.from_documents(docs, service_context=service_context)

async def event_generator(query: str):
    # Create a streaming chat completion
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": query}],
        stream=True
    )

    # Stream tokens as SSE
    async for chunk in response:
        if "choices" in chunk and len(chunk["choices"]) > 0:
            delta = chunk["choices"][0]["delta"]
            if "content" in delta:
                yield f"data: {delta['content']}\n\n"
    yield "data: [DONE]\n\n"

@app.get("/stream")
async def stream_response(request: Request, q: str):
    generator = event_generator(q)
    return StreamingResponse(generator, media_type="text/event-stream")

# To run: uvicorn this_file_name:app --reload

# Client side can connect to /stream?q=your+query and receive streamed tokens.