Hey guys, LD here, Spent months working on this. otherwise it would be free.
you don't just buy this node you buy everything i make, For future video gen models too.?
i'm also in sulphur discord.
H3 Studio 2 – interactive tour by Lora-Daddy
Thanks guys.
updated weekly.
workflow is a placeholder you do need the node pack unfortunately.
05/10/2026 Reference rail updated to allow 20 images 3 videos and 9 mp3
Does not use Ref mod. - get wardrobe clothing with no bodies to avoid transfer
Description
FAQ
Comments (9)
I feel WinAMP vibes
You are right, I have the same feeling. It Really Whips the Llama's Ass, yeah whips, not wipes.
yeahhh bro its 1 of the themes love it XD
@Lora_Daddy omg it actually is ! i just saw the drop down theme selection around 0:09 sec into the video ;)
just keep getting errors when trying to generate the prompt
1 of 1 clip(s) came back without their shot blocks - writing again…
⚠ the writer hit its token limit and was cut off - the end of the script is missing
⚠ missing field: detailed_description, overall_soundscape, non_diegetic_music
⚠ no [Shot N] blocks
My context length is set to 110K in LM studio. I have no problems when i use h3 Writer paired with ollama in other workflows....whats going on?
i had high hopes for this :/
FIXED: I had to change the Chat template of the quantized model in LM studio to {{ bos_token }}{% for message in messages %}{{ message['content'] }}{% endfor %}
Works now! Thank you!
don't need high context mines 32k what gpu you got, and what llm model?
I have a 4080, ive tried Qwen 3 12b, gemma 4 12b, gemma 4 26b, and qwen 3.8 27b. all have given that same error.
@Lora_Daddy FIXED: I had to change the Chat template of the quantized model in LM studio to {{ bos_token }}{% for message in messages %}{{ message['content'] }}{% endfor %}
Works now! Thank you!
Best workflow by far. Thanks for your work! Just three questions:
1.) Would it be possible to create an automated loop where you queue in 4 runs in advance? A fresh prompt each run for creating four unique videos.
2.) I keep having to restart ComfyUI whenever I want a new prompt generated after a video finished rendering. Otherwise it is stuck at generating a prompt. Why is that the case?
3) Which LLM model are you personally using exactly? Maybe mine is flawed resulting to 2).
Thanks!