Subtitle API

FastAPI service that fetches and generates subtitles for YouTube videos — backend for SubtitleHub.

Endpoints

Method	Path	Purpose
POST	/opensubtitles/login	OpenSubtitles auth (returns token)
GET	/opensubtitles	Search OpenSubtitles for movie/TV subs
GET	/download?url=<video>	Fetch captions as SRT/VTT/TXT
POST	/download	Same, POST variant
GET	/download/raw	Raw transcript text
GET	/playlist	Captions for a YouTube playlist
GET	/	Health check + endpoint list

How it works

Fetch strategy (fallback chain):

1. youtube-transcript-api — fast, direct caption extraction
2. yt-dlp — resilient pipeline for videos without auto-captions
3. YouTube timedtext proxy as last resort

Format conversion (format_output) → srt, vtt, txt, json.

Features

• Rate limiting: 30 req / 60s per IP (RATE_LIMIT_REQUESTS)
• CORS enabled for web clients
• Optional OpenSubtitles integration for movie/TV subtitle search

Run locally

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

Deploy

One-click on Render via render.yaml (uvicorn). If using OpenSubtitles, import OPENSUBTITLES_API_KEY as a Render secret — never commit it (this repo had one leaked in history; rotate it).

---

Built by Muhammad Zubair — Clinical psychologist → AI builder. Part of an AI-specialist portfolio.