ComfyUI, the node-based AI creation engine created by Yannik Marek, added native MiniMax H3 workflows on Aug. 2, turning the video model into editable graphs with local-runtime support. ComfyUI says the integration can run locally on Nvidia RTX 3060-class hardware.
The release adds the workflow layer around MiniMax's H3 weights, including editable templates and local-runtime support. RuntimeWire covered MiniMax's checkpoint release, 2K service path and community license separately. Here, the news is how ComfyUI has made the model inspectable and configurable inside its node-based engine.
Marek began ComfyUI in early 2023 as a solo Stable Diffusion experiment. ComfyUI says the project started when no available tool could chain two AI models into a repeatable workflow, prompting Marek to build and open-source his own node-based system. The resulting interface exposes models, parameters and processing stages on a graph rather than hiding them behind a single prompt box.
That structure makes rapid model integration central to ComfyUI's product. ComfyUI's announcement says H3 accepts text, images, video and audio. MiniMax's H3 model card says the model generates clips at 24 frames per second and up to about 15 seconds, with native stereo audio. ComfyUI packaged those capabilities into workflows for text-to-video, image-to-video, first-and-last-frame generation and reference-driven video. Reference media can carry a subject, motion or voice into a new clip, according to the ComfyUI announcement.
ComfyUI's H3 documentation requires version 0.30.0 or later and lists text-to-video, image-to-video and reference-to-video templates. Developers can inspect and modify each graph, replacing a model or processing stage without reconstructing the entire workflow.
Why ComfyUI's graph matters
ComfyUI says it optimized H3 for local use and paired the model with ready-made workflows instead of leaving developers to assemble the graph themselves. The announcement supports the RTX 3060 compatibility claim, but it does not establish comparable generation speed or output quality across GPUs.
Developers can use ComfyUI's H3 documentation to start from text-to-video, image-to-video and reference-to-video graphs. Because ComfyUI exposes the pipeline as nodes, a creator can alter one model, parameter or processing step, rerun the graph and preserve the rest of the process.
The local workflow has limits. MiniMax's official model card says H3-Base generates at 768p, while the documented full-2K process combines a locally deployed base model with hosted H3-Context-IR and H3-Regenerate-2K services. ComfyUI's tutorial describes a native canvas with a 768-pixel short edge, capped at 768 by 1,344 pixels.
MiniMax describes H3 as open weights, though its terms are more restrictive than an unrestricted open-source license. The MiniMax H3 Community License Agreement imposes geographic restrictions and requires separate written authorization when a commercial product or service using H3 generates more than $20 million in annual revenue.
ComfyUI's workflow bet
ComfyUI is building a common execution layer for visual models, preprocessing steps and post-production tools. ComfyUI said in April that it raised $30 million, bringing total funding to $47 million. TechCrunch reported a $500 million valuation for the round. Craft led the financing, with Pace Capital, Chemistry and TruArrow participating. ComfyUI reported 4 million users, more than 60,000 community-built nodes and over 150,000 daily downloads at the time.
The H3 integration shows how Marek's graph model can absorb a newly released checkpoint and turn it into a modifiable production process on the same day. MiniMax supplied the model. ComfyUI built the editable templates and local execution path needed to run it inside an existing creation engine.