Instagram Vision

FastAPI-based Instagram scraper built with instagrapi. It exposes endpoints for profiles, posts, comments, stories, reels, highlights, followers, followings and hashtag search. Authentication works via username/password, sessionid, or anonymously. Media can be downloaded through a dedicated endpoint. The service is ready for Railway deployment via Docker and integrates well with n8n via HTTP requests.

Visit the root path / to see a short welcome message. Use /login and /logout to manage sessions.

Quick start

pip install -r requirements.txt
uvicorn app.main:app --reload

Note that httpx is pinned to <0.28 in requirements.txt for compatibility with instagrapi.

Use the /login endpoint to obtain a token before calling other routes. The token must be supplied as ?token=... query parameter. Media content for a post or story can be retrieved through /download/{media_id}.

If you omit credentials when calling /login, an anonymous session is created. Some endpoints may require authentication to succeed.

n8n integration

The API can be accessed from n8n via the HTTP Request node. Provide the token as a query parameter and parse

the JSON response for further automation. Example cURL request:

curl -X POST -H "Content-Type: application/json" -d '{"username":"myuser","password":"mypass"}' \
  http://localhost:8000/login

Docker

Build and run:

docker build -t instagram-vision .
docker run -p 8000:8000 instagram-vision

Responsible use

Ensure you follow Instagram's Terms of Service and comply with local laws when scraping or downloading content.

License

Released under the MIT License.