<div data-youtube-video="https://youtu.be/SUJkkT6XLHI"><iframe src="https://www.youtube.com/embed/SUJkkT6XLHI" width="853" height="480" allowfullscreen></iframe></div> 
RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2086411696771522561?inviteCode=rh-v1111
<h2>MiniMax H3 Image-To-Video — 24 GB Edition</h2>
A MiniMax H3 video graph from episode 291, part of the six-times-acceleration series that compares model tier, LoRA, attention and cache settings.
<h2>Saved configuration</h2>
24 nodes / 29 links. One LoadImage input. UNETLoader <code>minimax_h3_fl2va_bf16.safetensors</code>; CLIPLoader <code>qwen3vl_32b_minimax_h3_bf16.safetensors</code> (type minimax); VAELoader video <code>minimax_h3_video_vae_fp16.safetensors</code> + audio <code>minimax_h3_audio_vae_fp32.safetensors</code>. LoRA <code>minimax_h3_turbo_v4_step600_ema_pruned_comfyui.safetensors</code> at 1. Sampling: <code>KSamplerSelect</code> euler + <code>BasicScheduler</code> beta, 6 steps, denoise 1.
<h2>How to use</h2>
This is the 24 GB general edition: bf16 weights, fp16 video VAE, one image and the turbo-v4 LoRA at strength 1. Six steps is the whole sampler budget — there is no 20-step fallback in this graph.
<h2>Limits and not-verified items</h2>
No inference was run for this listing: no speed, VRAM, quality, audio-sync or acceleration-factor claim is made. The "6x" in the episode title is the author's description and was not independently measured. Values are quoted from the saved nodes. H3 weights, LoRAs, encoder and VAEs are not bundled; check their licenses separately. The graph notes credit the official Comfy-Org/MiniMax-H3 model files; preserved as-is, no originality claim. The cover image is a promotional still, not a frame produced by this graph.
Tutorial: <a href="https://www.youtube.com/watch?v=SUJkkT6XLHI" rel="ugc">YouTube</a> · <a href="https://www.bilibili.com/video/BV17nuU6xE3g" rel="ugc">Bilibili</a>. Workflows supplied through AI-KSK; verify workflow, node, model, LoRA and cover rights separately. No blanket commercial or redistribution license is asserted.

MiniMax H3 Image-to-Video 6x Speedup - 24 GB Edition

<div data-youtube-video="https://youtu.be/SUJkkT6XLHI"></div> 
RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2086411696771522561?inviteCode=rh-v1111
MiniMax H3 Image-To-Video — 24 GB Edition
A MiniMax H3 video graph from episode 291, part of the six-times-acceleration series that compares model tier, LoRA, attention and cache settings.
Saved configuration
24 nodes / 29 links. One LoadImage input. UNETLoader <code>minimax_h3_fl2va_bf16.safetensors</code>; CLIPLoader <code>qwen3vl_32b_minimax_h3_bf16.safetensors</code> (type minimax); VAELoader video <code>minimax_h3_video_vae_fp16.safetensors</code> + audio <code>minimax_h3_audio_vae_fp32.safetensors</code>. LoRA <code>minimax_h3_turbo_v4_step600_ema_pruned_comfyui.safetensors</code> at 1. Sampling: <code>KSamplerSelect</code> euler + <code>BasicScheduler</code> beta, 6 steps, denoise 1.
How to use
This is the 24 GB general edition: bf16 weights, fp16 video VAE, one image and the turbo-v4 LoRA at strength 1. Six steps is the whole sampler budget — there is no 20-step fallback in this graph.
Limits and not-verified items
No inference was run for this listing: no speed, VRAM, quality, audio-sync or acceleration-factor claim is made. The "6x" in the episode title is the author's description and was not independently measured. Values are quoted from the saved nodes. H3 weights, LoRAs, encoder and VAEs are not bundled; check their licenses separately. The graph notes credit the official Comfy-Org/MiniMax-H3 model files; preserved as-is, no originality claim. The cover image is a promotional still, not a frame produced by this graph.
Tutorial: <a href="https://www.youtube.com/watch?v=SUJkkT6XLHI" rel="ugc">YouTube</a> · <a href="https://www.bilibili.com/video/BV17nuU6xE3g" rel="ugc">Bilibili</a>. Workflows supplied through AI-KSK; verify workflow, node, model, LoRA and cover rights separately. No blanket commercial or redistribution license is asserted.

01__MiniMax_H3____6___24GB___.json

v1.0

https://youtu.be/SUJkkT6XLHI 
 
This workflow is the 24GB general image-to-video acceleration version for MiniMax H3. It is prepared for users who want a practical 6x-speed style workflow that can run on more accessible local or cloud GPU configurations. 
 
The graph uses the MiniMax H3 image-to-video route with Qwen3-VL text encoding, MiniMax H3 video and audio VAE components, turbo acceleration, and a connected SaveVideo output. The 24GB label indicates the lighter general-purpose hardware target in this group. 
 
This package is meant for practical ComfyUI and RunningHub users who want a ready-made MiniMax H3 graph instead of rebuilding loader, encoder, VAE, sampler, and video export connections by hand. The workflow is especially useful for speed comparison, hardware planning, and repeatable prompt tests. Keep the same source image and prompt when comparing versions so the difference you see comes from the workflow route, not from changing creative variables. 
 
Main features: 
 
- MiniMax H3 image-to-video workflow for 24GB-class use 
- 6x-speed acceleration concept for faster iteration 
- Single image input drives the starting frame and subject identity 
- Qwen3-VL MiniMax text encoding with MiniMax H3 VAE components 
- 16:9 video route for common horizontal output 
- Turbo-style short-step sampling path 
- Good starter version before moving to 48GB or 80GB variants 
 
Suggested workflow: 
 
Begin with the 24GB version if you want the most practical compatibility target. Use a clean input frame, keep the subject identity explicit, and ask for controlled motion first. Once the motion and lip sync are stable, you can move to the higher-memory variants for comparison. 
 
For better comparisons, make one small test first, review identity stability and temporal motion, then raise prompt complexity only after the base route behaves correctly. When using speaking-character prompts, keep the face visible, avoid too many fast cuts, and separate action, expression, and camera instructions clearly. 
 
RunningHub Workflow 
 
Try the workflow online right now - no installation required. 
Workflow: https://www.runninghub.ai/post/2086424102268678145?inviteCode=rh-v1111 
 
If the results meet your expectations, you can later deploy it locally for customization. 
 
Fan Benefits: Register to get 1000 points + daily login 100 points - enjoy 4090 performance and 48 GB super power! 
 
Bilibili Updates (Mainland China &amp; Asia-Pacific) 
 
If you're in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown. 
Bilibili Video: https://www.bilibili.com/video/BV17nuU6xE3g/ 
 
Support Me on Ko-fi 
 
If you find my content helpful and want to support future creations, you can buy me a coffee. 
Every bit of support helps me keep creating. 
Ko-fi: https://ko-fi.com/aiksk 
 
Business Contact 
 
For collaboration or inquiries, please contact aiksk95 on WeChat. 
 
打开下方链接即可在线体验，无需安装。 
工作流：https://www.runninghub.ai/post/2086424102268678145?inviteCode=rh-v1111 
 
如果你觉得效果理想，也可以在本地进行自定义部署。 
 
粉丝福利：注册即可领取 1000 积分，每日登录再领 100 积分，体验 4090 和 48GB 大显存性能。 
 
B站视频（中国大陆及亚太地区） 
 
如果你在中国大陆或亚太地区，可以通过下面的视频查看工作流的实测效果与创作思路。 
B站视频：https://www.bilibili.com/video/BV17nuU6xE3g/ 
 
我会在夸克网盘持续更新模型资源： 
https://pan.quark.cn/s/07bdc81784ce

Files with SHA256: 803fa88586ee0bd62eca20d95bc36569f65665a9e083ad99303e2deeda936470