# LLM video animation questions. Why can’t I get the results ? — retained record

> Retained reddit record from reddit:generativeAI.

Canonical: https://highsignal.app/data/records/3f990bf4a6d92434

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:generativeAI)  published 2026-10-23 00:39 ingested 2026-10-24 02:39

[original source ↗](https://www.reddit.com/r/generativeAI/comments/1xczps6/llm_video_animation_questions_why_cant_i_get_the/)

## retained text



```
So I’ve recently began experimenting with video animation using Claude, codex, even grok. 

Bear with me if something I say is off I’m not a CS major.

Basically the idea of anyone is unfamiliar, you have your agent write code that is in like html or JS or TS. 

The agent makes the scenes in code (aka draws them) 

Then you use either hyper frames or remotion to render the video. 

I went down a deep adhd rabbit hole for 20 hours trying to recreate some YouTube channel animations. 

The closest I got was on Claude.  
I probably experimented with 10+ different skills. 

But ultimately I was never satisfied. I couldn’t get any model to basically replicate the quality of a particular video I saw, even though my reference videos were for sure made this way. 

For example in the linked Instagram video the creator has a nice reel explaining how AirPods work. The animation style is great. I fail to reproduce results like this . 

I think Claude is capable but I just don’t have the work flow or prompting. 

If anyone has any advice please help. 💕💕💕💕
```



document fetched 2026-10-24 00:39 · canonical https://reddit.com/r/generativeAI/comments/1xczps6/llm_video_animation_questions_why_cant_i_get_the
