joeygambino

/

joyai-echo-ltx25-echoVid-comfy-native

like 7

I have a system prompt built into the LLM Enhance node already, so as long as you choose one of the long_story, short_story or revise options in that node, and you just write out your ideas in the SCENE IDEA box, it will take care of the rest. I will say that depending which models you use will depend on how well the prompt is structured. So far I have used GLM 5.2 (good structure), minimax-m3 (reasoning model so it takes a long time, but does pretty good, and it's great at describing characters), deepseek4pro is pretty good too. Works for local models too, Qwen3.6 and Gemma4 both do well. Mileage may vary.

Where to find it??