this is a simple canvas animation that i had a local LLM put together.

the quality of the code that these models can quickly spit out has dramatically improved recently.

the model: Qwen3.6-35B-A3B-Uncensored (8-Bit quant)

*note: i am publishing this again to test a fix i just made to my publish script. i also made the cube smaller, to add a little variety.