Pika H3 Flex ComfyUI Workflow | 8-Step Ref2Video
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
ComfyUI model file · LoRA (step distillation)
MiniMax H3 Turbo 4-step LoRA (ComfyUI) · 4-step speed LoRA for MiniMax H3
This filename shows up in community MiniMax H3 setup guides as the Turbo LoRA that cuts sampling to 4 steps. It is a renamed copy of larryvrh’s MiniMax H3 Turbo 4-step LoRA converted for ComfyUI. No public host lists a file under exactly this name, so download a verified copy below.
We could not find a hash for this exact filename. The sources below are the original Turbo LoRA and its ComfyUI conversion, with published hashes.
minimax_h3_turbo_4steps_comfyui.safetensors
The short version
larryvrh distilled MiniMax H3 into a LoRA that produces usable video in 4 sampling steps. ComfyUI needs the LoRA keys renamed to its own format, so several people published converted copies with “comfyui” in the name. minimax_h3_turbo_4steps_comfyui.safetensors is one such name, used in setup guides that mirror files on ModelScope.
Because no host lists this exact file with a hash, we point to the conversion with a published SHA256 (drbaph’s minimax_h3_turbo_4step_pruned_comfyui) and the original repo. One guide notes its copy is meant for the full (unpruned) H3 model; the pruned conversion matches the pruned models most workflows use.
Install
larryvrh’s 4-step LoRA converted for ComfyUI, 620 MB. Rename if your workflow needs it.
saves as minimax_h3_turbo_4step_pruned_comfyui.safetensors
Original Turbo LoRA and its checkpoints.
saves as minimax_h3_turbo_4step.safetensors
ComfyUI/ └─ models/ ├─ … └─ loras/ └─ minimax_h3_turbo_4steps_comfyui.safetensors
$ cd ComfyUI/models/loras$ wget -O minimax_h3_turbo_4steps_comfyui.safetensors \$ https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI/resolve/main/minimax_h3_turbo_4step_pruned_comfyui.safetensors
Safetensors files hold tensors only and cannot run code. The source does not publish a hash, so compare the file size after downloading.
Use
LoraLoaderModelOnly patches the base model with this LoRA. Put it between the model loader and the sampler.
MODEL
Works with
minimax_h3_turbo_4steps_comfyui.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 |
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.
19 workflows
These workflows load another fast sampling model in the same folder, or they are fast sampling workflows that already have a LoraLoaderModelOnly node. Select minimax_h3_turbo_4steps_comfyui.safetensors there.
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_taomate_3step_lora_avg_rank_19_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_fl2v_turbo_8step_v1.0_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_4step_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_4step_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_4step_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_4step_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_v4_step600_ema_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_fl2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_fl2v_turbo_8step_v1.0_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_fl2va_4step_lora.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
LoraLoaderModelOnly
minimax_h3_turbo_4step_pruned_comfyui.safetensors
Replace with minimax_h3_turbo_4steps_comfyui.safetensors
Compare
| File | Relation | Size | Best for |
|---|---|---|---|
| minimax_h3_turbo_4step_pruned_comfyui.safetensors | ComfyUI conversion · for pruned H3 | 620 MB | Pruned FL2VA/Ref2VA models |
| minimax_h3_turbo_4step.safetensors | Original · larryvrh | 780 MB | Source weights |
| minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors | Official alternative · lightx2v via Comfy-Org | 1.96 GB | FL2VA at 768p, 4 steps |
Troubleshoot
Why: The workflow asks for this exact filename, but ComfyUI cannot find it in models/loras. Usually the file is missing, sits in a different folder, or was saved under another name such as minimax_h3_turbo_4STEPS_comfyui.safetensors.
Fix: Move the file into ComfyUI/models/loras/, rename it to match the workflow (or pick it again in the dropdown), then refresh the node list or 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 the full size listed by the source. Download it again from Hugging Face · drbaph (ComfyUI conversion) if not.
Rather skip the setup?
Run Pika H3 Flex ComfyUI Workflow in your browser on a cloud GPU. Nothing to download.
FAQ
ComfyUI/models/loras/. Then refresh the node list or restart ComfyUI so LoraLoaderModelOnly lists the file.Sources: larryvrh/MiniMax-H3-Turbo-Lora · drbaph/MiniMax-H3-Turbo-Lora-ComfyUI
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.