SamurAIGPT/Text-To-Video-AI

Generate video from text using AI

What it solves

This project provides an automated pipeline to transform a simple text topic into a complete video, eliminating the need for manual scriptwriting, voiceover recording, and footage gathering. It is specifically designed for creators making short-form content like YouTube Shorts, Instagram Reels, and TikToks.

How it works

The system orchestrates several AI models to build a video end-to-end:

  1. Scripting: Uses LLMs (OpenAI, Groq, or Google Gemini) to generate a script based on the user's topic.
  2. Voiceover: Converts the script to speech using EdgeTTS or ElevenLabs.
  3. Visuals: Fetches relevant B-roll footage from Pexels or generates AI video segments via the Muapi API (supporting models like Google Veo, Sora, and Kling).
  4. Captions: Uses Whisper or Deepgram to create timestamped captions that are synchronized with the audio.
  5. Composition: Combines these elements into a final video file using the React/Remotion composition engine.

Who it’s for

Content creators, educators, and social media managers who want to produce high-quality, narrated videos quickly without professional editing software.

Highlights

  • Flexible AI Providers: Support for multiple LLM, TTS, and STT providers.
  • AI Video Integration: Integration with Muapi to access cutting-edge text-to-video models.
  • Customizable Styling: Full control over caption fonts, colors, and video orientation (portrait or landscape).
  • Multiple Deployment Options: Can be run locally, via Google Colab, or through a managed Premium API.