Agents365-ai/video-podcast-maker

What it solves

It automates the complex process of turning a simple topic into a professional 4K video podcast. It eliminates the need for manual video editing and technical setup by coordinating research, scriptwriting, voice synthesis, and visual rendering into a single pipeline.

How it works

The project functions as a pipeline that can be driven by a coding agent. It follows these steps:

1. Research & Scripting: Generates research notes and a narration script based on a provided topic.
2. Audio Synthesis: Uses local TTS backends (Microsoft Edge for free or Azure for premium) to convert the script into speech.
3. Visual Composition: Uses Remotion (a React-based video framework) to compose the video, including 4K subtitles and animations.
4. Asset Integration: Pulls in B-roll, stills, and overlays from various sources (user files, stock libraries, or AI generators).
5. Final Mixing: Uses FFmpeg to mix the rendered video with background music and export the final file.

Who it’s for

Content creators who want to produce high-quality video podcasts for platforms like YouTube, Bilibili, Douyin, Xiaohongshu, and WeChat Channels without needing deep technical skills in video editing or coding.

Highlights

• End-to-End Automation: Goes from a text topic to a final 4K MP4 file.
• Multi-Platform Support: Tailored outputs for various social media platforms and multi-language support (English and Chinese).
• Flexible TTS: Built-in support for free Edge TTS and professional Azure Speech.
• AI-Driven Assets: Integration with AI tools for generating B-roll and thumbnails.
• Vertical Shorts: Ability to generate 9:16 highlight clips from longer videos.