SamurAIGPT/Text-To-Video-AI — reverse-engineered prompt

Build me a simple text to video AI tool that turns one topic or prompt into a finished social media video. I want it to write an engaging short script, create a natural voiceover, pull relevant background clips from Pexels, add timed captions, and export a finished MP4.

It should be easy to run from the command line with something like python app.py "my topic", and save the result as rendered_video.mp4. Please support portrait videos for TikTok, Reels, and YouTube Shorts, plus landscape videos too. Let me choose OpenAI, Groq, or Gemini for writing, Edge TTS or ElevenLabs for voice, and Whisper or Deepgram for caption timing through a .env file.

Also include clear setup instructions, an example env file, and sensible defaults so a beginner can get it working locally or in Colab. If you need current docs for any API, look them up online.