MiniMax H3: ComfyUI Workflow Examples

• Text-to-Video (T2V): Generate videos from text prompts
• Image-to-Video (I2V): Generate videos from an input image, with optional first/last-frame control
• Reference-to-Video (R2V): Generate videos that lock in a character, style, motion, camera move, or voice from reference images, videos, and audio

​

Key features

• Native stereo audio: Dialogue, sound effects, and music are generated together with the video, synced in one MP4
• Multimodal context: Text, images, video, and audio references can be combined in one generation
• Reference-driven generation: Lock a character’s identity, a style, a motion, a camera move, or a voice from reference materials
• Instruction following: Describe the relationship between references and the target shot in natural language
• Accurate text rendering: Spelled-out text and brand elements render cleanly
• Open weights: Run locally in ComfyUI with full control over every parameter

​

Getting started

1. Update ComfyUI to version 0.30.0 or later
2. Go to Template Library > Video > choose any MiniMax H3 workflow
3. Follow the pop-up to download models and run the workflow

​

Setting the output resolution

• Aspect ratio: Pick a preset such as 16:9 (Widescreen), 9:16 (Portrait Widescreen), or 1:1 (Square)
• Megapixels: Target total pixel count for the output. Higher values give larger frames; lower values run faster
• Multiple: The computed resolution is rounded to the nearest multiple of this number. Keep it at 32 to match H3’s resolution grid

​

ComfyUI Native Workflows

​

MiniMax H3 Text to Video (T2V)

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── minimax_h3_fl2va_pruned_int8_convrot.safetensors
│   ├── 📂 text_encoders/
│   │   └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors
│   └── 📂 vae/
│       ├── minimax_h3_video_vae_fp16.safetensors
│       └── minimax_h3_audio_vae_fp32.safetensors

​

Prompting tips

1. Describe the whole scene: State the overall scene first (location, character, what is happening), then break it into timed shots
2. Shots, camera, and audio: Describe the shots, camera moves, and the accompanying audio (dialogue, SFX, music) in one prompt block
3. Resolution: H3’s native canvas is a 768px short edge, capped at 768x1344 pixels and rounded to a multiple of 32
4. Duration: The duration input snaps to the model’s 17-frame-per-block (17k+5) grid at 24fps
5. First/last frame: Connect images to first_frame and/or last_frame on the MiniMaxH3ImageToVideo node to turn this workflow into first/last-frame image-to-video

​

Prompt writing guide

​

MiniMax H3 Image to Video (I2V)

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── minimax_h3_fl2va_pruned_int8_convrot.safetensors
│   ├── 📂 text_encoders/
│   │   └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors
│   └── 📂 vae/
│       ├── minimax_h3_video_vae_fp16.safetensors
│       └── minimax_h3_audio_vae_fp32.safetensors

​

Prompting tips

1. Keyframes: The first_frame and last_frame inputs are optional; the model generates the motion between them
2. Prompt: Describe the shots, motion, and the accompanying audio (dialogue, SFX, music) in one block
3. Resolution: H3’s native canvas is a 768px short edge, capped at 768x1344 pixels and rounded to a multiple of 32
4. Duration: The duration input snaps to the model’s 17-frame-per-block (17k+5) grid at 24fps

​

Prompt writing guide

​

MiniMax H3 Reference to Video (R2V)

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── minimax_h3_ref2va_pruned_int8_convrot.safetensors
│   ├── 📂 text_encoders/
│   │   └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors
│   └── 📂 vae/
│       ├── minimax_h3_video_vae_fp16.safetensors
│       └── minimax_h3_audio_vae_fp32.safetensors

​

Prompting tips

1. Reference by tag: Reference each input by tag in the exact order it was connected, for example <Picture 1>, <Video 1>, <Audio 1>
2. Assign each reference a job: State which reference drives which part of the shot (identity, style, motion, camera, voice). Explicit assignments tend to work much better
3. Limits: Up to 9 reference images, 3 reference videos (each can carry its own soundtrack), and 3 standalone reference audio clips
4. ref_image_size: match scales references down to the generation resolution for speed; max keeps up to a 2048px short edge for stronger identity fidelity at the cost of speed
5. Note: R2V uses the ref2va diffusion model, a different set of weights from the fl2va model used by the T2V and I2V workflows

​

Prompt writing guide

​

Speeding up generation with Sage Attention

1. Install the sageattention Python package. Download the wheel that matches your PyTorch and CUDA versions from the SageAttention releases page, then install it with pip install <wheel-file>.
2. Install the KJNodes custom nodes, which provide the Patch Sage Attention KJ node. Use the ComfyUI Manager, or clone the repository into ComfyUI/custom_nodes/ and restart ComfyUI.
3. Add a Patch Sage Attention KJ node to the workflow and connect it between the UNETLoader and the BasicGuider node: its model input receives the model from the UNETLoader, and its model output feeds the model input of the BasicGuider. Set sage_attention to auto.
4. Run the workflow as usual. Only the guider needs the patch; the scheduler only generates the sigmas and can stay as is.

• Sage Attention requires float16 or bfloat16 tensors. MiniMax H3 runs some layers in other dtypes, so you may see “Input tensors must be in dtype of torch.float16 or torch.bfloat16, using pytorch attention instead” messages in the console. These are expected; the affected layers fall back to standard attention and generation still works.
• Alternatively, you can enable Sage Attention globally by launching ComfyUI with the --use-sage-attention flag instead of adding the node.