Laravel + Python FastAPI: Image OCR Demo

Integrating Laravel with Python FastAPI for Image OCR

Modern web applications increasingly need to extract text from images — receipts, scanned documents, ID cards, and more. Laravel is excellent at handling HTTP requests, queuing jobs, and managing user-facing logic, but Python's ecosystem dominates when it comes to OCR libraries. The solution? Run a Python OCR service as a standalone API using FastAPI, then call it from Laravel over HTTP.

This tutorial from Laravel Daily demonstrates exactly that pattern in a focused 22-minute video, complete with a GitHub repository you can clone and explore.

Why FastAPI as the OCR Bridge?

FastAPI is a modern, high-performance Python web framework that makes it trivial to expose any Python function — including OCR processing — as a REST endpoint. Key reasons to choose it for this pattern:

• Speed: FastAPI is built on Starlette and uses async I/O, making it fast enough for real-time image processing requests.
• Auto-generated docs: Swagger UI is available out of the box, so you can test your OCR endpoint before wiring up Laravel.
• Simple deployment: The FastAPI service can run as a separate process or container alongside your Laravel app.

How Laravel Calls the FastAPI Service

The general flow looks like this:

1. A user uploads an image through the Laravel frontend.
2. Laravel stores the image and dispatches an HTTP request to the FastAPI service.
3. FastAPI receives the image, runs the OCR script (e.g., using Tesseract or a similar library), and returns the extracted text as JSON.
4. Laravel receives the response and stores or displays the result.

A minimal Laravel HTTP call to the FastAPI endpoint might look like this:

use Illuminate\Support\Facades\Http;

$response = Http::attach(
    'file',
    file_get_contents($imagePath),
    'image.png'
)->post('http://localhost:8000/ocr');

$extractedText = $response->json('text');

On the Python side, a minimal FastAPI endpoint that accepts the uploaded file and runs OCR could be structured as:

from fastapi import FastAPI, UploadFile, File
import pytesseract
from PIL import Image
import io

app = FastAPI()

@app.post("/ocr")
async def run_ocr(file: UploadFile = File(...)):
    contents = await file.read()
    image = Image.open(io.BytesIO(contents))
    text = pytesseract.image_to_string(image)
    return {"text": text}

Note: The exact implementation details are covered in the premium video. The snippets above illustrate the general pattern.

Architecture Considerations

When running two separate services (Laravel + FastAPI), keep these points in mind:

• Networking: Both services need to be able to reach each other. In a Docker Compose setup, use service names as hostnames.
• Error handling: Always handle cases where the FastAPI service is unavailable — wrap your Http::post() call in a try/catch and check $response->successful().
• Security: If the FastAPI service is internal only, bind it to localhost or a private network and never expose it publicly without authentication.
• Timeouts: OCR on large images can be slow. Set an appropriate timeout on the Laravel Http client: Http::timeout(30)->attach(...).

Key Takeaways

• Laravel's Http facade makes it straightforward to send multipart file uploads to any external API.
• FastAPI is a lightweight, production-ready choice for wrapping Python ML or OCR scripts as HTTP services.
• Separating concerns — Laravel handles the web layer, Python handles the heavy processing — keeps both codebases clean.
• The pattern is reusable: swap OCR for any other Python capability (image classification, NLP, PDF parsing).
• A repository is included with the tutorial, giving you a working starting point.

Full tutorial (video + repository) available at: https://laraveldaily.com/post/laravel-python-fastapi-image-ocr-demo