video1.0<iframe src="https://www.loom.com/embed/e3650469f3094e0990b858c1d3968466" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/e3650469f3094e0990b858c1d3968466-6420bac11add53ce.gif593.068This Loom demonstrates how the AI video generator builds a complete video from a script, audio, and visuals. It starts with an LLM-based horror story brainstorming step, then generates a script, converts it to speech using local TTS options such as Kokoro, and optionally adds background music from locally uploaded files. For visuals, it explains three approaches: AI-generated images, stock media scraped from sources like Pexels or Pinterest, or a local folder of images and videos with filters such as dynamic zoom and fade in fade out. The speaker notes an example run where 2,070 frames were added out of about 4,400 total before the final video was produced, with caveats that stock visuals are less accurate than AI or local sources.