MiniMax H3 20 Step Quality Baseline Ref2VA

AI-KSK ComfyUI workflow: MiniMax H3 20 Step Quality Baseline Ref2VA.

This workflow is the 20-step quality baseline for reference-to-video. It runs the MiniMax H3 Ref2VA INT8 model with the Qwen3-VL 32B text encoder on a 20-step simple schedule, with no turbo LoRA, so the result represents the full-strength reference of the whole comparison series.

Main features:
- 20-step simple schedule as the quality baseline
- MiniMax H3 Ref2VA INT8 model with Qwen3-VL 32B encoder
- No turbo LoRA: full-strength model behavior
- Reference image plus text prompt input
- Video and audio VAE decode to saved MP4 output

Suggested workflow:
Use this route to judge what the model can actually do before tuning. Compare every accelerated variant in the series against this baseline with the same image and prompt to see exactly what each speed setting costs.

RunningHub workflow (try online, no install needed):
https://www.runninghub.ai/zh-cn/post/2098325682601345025?inviteCode=rh-v1111

AI-KSK RunningHub workflow hub:
https://www.runninghub.ai/user-center/1892566146468511746?inviteCode=rh-v1111

Bilibili tutorial:
https://www.bilibili.com/video/BV16DYL6FEsh/

RunningHub registration and AI tools:
https://www.runninghub.ai/?inviteCode=rh-v1111
https://www.runninghub.ai/ai-generator/2046514150500524034?inviteCode=rh-v1111
https://www.runninghub.ai/vip-rights/3?inviteCode=rh-v1111
https://rhtv.runninghub.ai/projects?inviteCode=rh-v1111

Stable global access:
https://lelians.com/#/register?code=mFIJ5tZn

TokenRhythm AI model credits:
https://tokenrhythm.studio/i/rf_tr_kM6wekDquDDZQD4PUNmgiuHA

Youyun Cloud GPU:
https://passport.compshare.cn/register?referral_code=A7fTQUYY9lCDFJwA6MMvoq
https://www.notion.so/2b72459a3ac6808f8282f2c687ed8f52

RunPod Cloud GPU:
https://runpod.io/?ref=sxk5n5m7

Support AI-KSK:
https://ko-fi.com/aiksk