MiniMax H3 Long Video Generation | Motion Context V2
Model Preview Override
taeh3.safetensors
Replace with taeh3_2d_kijai.safetensors
ComfyUI model file · Tiny VAE (preview decoder)
TAEH3 2D by Kijai · 2D preview decoder for MiniMax H3
taeh3_2d_kijai.safetensors is Kijai’s 2D tiny VAE for MiniMax H3 previews, renamed so it does not overwrite madebyollin’s taeh3.safetensors. Put it in ComfyUI/models/vae_approx and select it in your preview node.
Kijai’s own repo ships this file as vae_approx/taeh3.safetensors. Rename it to taeh3_2d_kijai.safetensors after downloading, or use the pre-renamed copy below.
taeh3_2d_kijai.safetensors
The short version
Kijai trained this small 2D decoder soon after MiniMax H3 came out, so ComfyUI could show real-color previews instead of Latent2RGB smudges. It decodes each latent frame on its own, so previews look like a slideshow of stills rather than smooth motion.
Its original name, taeh3.safetensors, clashes with madebyollin’s later temporal TAEH3, a different and better model. t8star’s Taeh3-Comfy pack renamed Kijai’s file to taeh3_2d_kijai.safetensors without changing the tensors, and that name is what most guides now use.
Install
Pre-renamed copy, tensors unchanged.
saves as taeh3_2d_kijai.safetensors
Original upload. Rename after download.
saves as taeh3.safetensors
ComfyUI/ └─ models/ ├─ … └─ vae_approx/ └─ taeh3_2d_kijai.safetensors
$ cd ComfyUI/models/vae_approx$ wget -O taeh3_2d_kijai.safetensors \$ https://huggingface.co/t8star/Taeh3-Comfy/resolve/main/models/vae_approx/taeh3_2d_kijai.safetensors$ sha256sum taeh3_2d_kijai.safetensors# f0f60fa072089997f817402098c2fd90777cb2660dd79cf5df42fc1e3e08e527
Safetensors files hold tensors only and cannot run code. Still compare the SHA256: a failed download often leaves an HTML error page saved under the model’s filename.
Use
TAEH3 does not touch your final video. It only decodes sampler previews. Model Preview Override from KJNodes reads it from models/vae_approx and patches the MODEL before your sampler.
Model Preview Override comes from the ComfyUI-KJNodes custom node pack. Install it with ComfyUI Manager first.
MODEL
tiny_vaetaeh3_2d_kijai.safetensors
ModelPreviewOverrideKJ
model●
previewlive RGB
Turn previews on
Set ComfyUI’s preview method to auto or TAESD so the sampler shows them.
Final output unchanged
The full video VAE still decodes the result. TAEH3 only speeds up previews.
Works with
taeh3_2d_kijai.safetensors is one part of the stack. A workflow that reports a missing model usually needs these too. Use the Ref2VA diffusion model instead of FL2VA for reference-to-video workflows.
| File | Role | Folder | Size | Download |
|---|---|---|---|---|
| minimax_h3_fl2va_pruned_int8_convrot.safetensors | Diffusion model (FL2VA) | models/diffusion_models/ | 20.97 GB | Get |
| qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors | Text encoder (Qwen3-VL 32B) | models/text_encoders/ | 15.69 GB | Get |
| minimax_h3_video_vae_fp16.safetensors | Video VAE | models/vae/ | 5.21 GB | Get |
| minimax_h3_audio_vae_fp32.safetensors | Audio VAE | models/vae/ | 605 MB | Get |
| minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors | Turbo LoRA, 4 steps(optional) | models/loras/ | 1.96 GB | Get |
Run it
We match workflows by the model files they load: this exact file first, then other versions of it, then workflows built for the same use, then other models from the same family.
2 workflows
These workflows load another MiniMax H3 model. Select taeh3_2d_kijai.safetensors in their Model Preview Override node to use it instead.
Model Preview Override
taeh3.safetensors
Replace with taeh3_2d_kijai.safetensors
Model Preview Override
taeh3.safetensors
Replace with taeh3_2d_kijai.safetensors
Compare
| File | Relation | Size | Best for |
|---|---|---|---|
| taeh3_2d_kijai.safetensors | This page · Kijai 2D, renamed | 9.8 MB | Light per-frame previews |
| taeh3.safetensors | Different model · madebyollin temporal | 22.7 MB | Smooth motion previews; recommended |
Troubleshoot
Why: The workflow asks for this exact filename, but ComfyUI cannot find it in models/vae_approx. Usually the file is missing, sits in a different folder, or was saved under another name such as taeh3.safetensors (Kijai/MiniMax-H3-TAE).
Fix: Move the file into ComfyUI/models/vae_approx/, rename it to match the workflow (or pick it again in the dropdown), then refresh the node list or restart ComfyUI.
Why: Model Preview Override is not built into ComfyUI. It comes from the ComfyUI-KJNodes custom node pack.
Fix: Install ComfyUI-KJNodes with ComfyUI Manager (or git clone it into ComfyUI/custom_nodes), then restart ComfyUI.
Why: The download is incomplete, or the browser saved an HTML sign-in or error page under the model’s filename.
Fix: Check that the file is exactly 9,791,388 bytes and that its SHA256 matches the value above. Download it again from Hugging Face · t8star/Taeh3-Comfy if not.
Rather skip the setup?
Run MiniMax H3 Long Video Generation in your browser on a cloud GPU. Nothing to download.
FAQ
f0f60fa072089997f817402098c2fd90777cb2660dd79cf5df42fc1e3e08e527), but not the same as madebyollin’s taeh3.safetensors. Never overwrite one with the other.ComfyUI/models/vae_approx/. Then refresh the node list or restart ComfyUI so Model Preview Override lists the file.Sources: Kijai/MiniMax-H3-TAE · t8star/Taeh3-Comfy
RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.