MiniMax FastH3 4-Step ComfyUI Workflow | MiniMax H3 Super Fast T2V and I2V Generator - v1.0 Showcase
MiniMax FastH3 four-step video drafts: Fast cinematic clips from text or images.
Who it's for: creators who want this pipeline in ComfyUI without assembling nodes from scratch. Not for: one-click results with zero tuning - you still choose inputs, prompts, and settings.
Open preloaded workflow on RunComfy
Open preloaded workflow on RunComfy (browser)
Why RunComfy first
- Fewer missing-node surprises - run the graph in a managed environment before you mirror it locally.
- Quick GPU tryout - useful if your local VRAM or install time is the bottleneck.
- Matches the published JSON - the zip follows the same runnable workflow you can open on RunComfy.
When downloading for local ComfyUI makes sense - you want full control over models on disk, batch scripting, or offline runs.
How to use (local ComfyUI)
1. Load inputs (images/video/audio) in the marked loader nodes.
2. Set prompts, resolution, and seeds; start with a short test run.
3. Export from the Save / Write nodes shown in the graph.
Expectations - First run may pull large weights; cloud runs may require a free RunComfy account.
Overview
Generate MiniMax H3 videos dramatically faster on RunComfy. Turn text or images into cinematic videos with prompt-controlled motion, atmosphere, and sound, powered by the optimized FastH3 workflow for much faster generation.
Important nodes:
Key nodes in Comfyui FastH3 4-Step ComfyUI Workflow
MiniMaxH3ImageToVideo (#219)
Central conditioning hub for both T2V and first-frame I2V. Provide your prompt and timing inputs, and optionally connect first_frame for continuity-sensitive shots. Adjust width and height for scale and aspect while keeping them as multiples of 32. Length comes from the duration-to-frames logic so you rarely need to set it manually.
MiniMaxH3SigmaShift (#6007)
Applies a sigma shift tailored to MiniMax H3’s FL2VA schedule so the short-path sampler remains stable. Leave it as configured for the FastH3 route; it is primarily a compatibility and quality safeguard. Consider changing only if you swap H3 weights or schedulers.
BasicScheduler (#212)
Provides an 8-step schedule that pairs well with the 4-step weight path for a practical quality baseline. You can reduce steps for faster drafts or lift them slightly for more stability, noting the trade-off between detail and speed. Step count interacts with sigma shift and sampler choice, so make changes thoughtfully.
SamplerCustomAdvanced (#213) with KSamplerSelect (#211)
Runs the short denoising trajectory used by FastH3. The selected sampler is tuned for multi-step efficiency and coherence with very few updates. Keep this pairing unless you are experimenting with alternative samplers validated for H3.
ResolutionSelector (#115)
High-level control for aspect and pixel budget. Use it to scale the look without touching model internals, and prefer resolutions that translate to about 0.7–1.0 megapixels for very fast drafts, then scale up when locking final shots. Outputs are quantized to keep decoders efficient.
VHS_VideoCombine (#6104)
Combines decoded frames and audio into a single MP4. Quality and size can be balanced via its compression and frame rate controls. It is part of Video Helper Suite, which you can explore here: Kosinkadink/ComfyUI-VideoHelperSuite.
Notes
MiniMax FastH3 4-Step ComfyUI Workflow | MiniMax H3 Super Fast T2V and I2V Generator - see RunComfy page for the latest node requirements.