Most of what earned @ComfyUI a $500M valuation today was built by people we don't employ.
We just closed $30M, led by @craft_ventures with @PaceCap, @chemistry, and Tru Arrow.
Total: $47M.
Where it goes:
→ Comfy Cloud
→ Collaborative workflows
→ Comfy Local
→ Ecosystem stability
→ Day-one model support
We're not building a walled garden. We're building open infrastructure, the kind that lasts and empowers rather than locks in.
This started as a community project. 4M users and $47M later, it still is.
If you haven't built something in ComfyUI yet, come find us: comfy.org/
It handles the prep you'd normally do by hand: cutting, cropping, format conversion, resizing, compression, frame extraction, speed changes, background removal, and local RIFE interpolation.
He's not a professional software engineer. He just kept fixing his own workflow and turned it into something everyone can use. FrameShift takes care of the prep before you even open the graph.
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
We don't have a recruiter combing LinkedIn for us. I just pay attention to what people ship, and Willie was contributing to @ComfyUI long before any of this. His contact-sheet prompting technique is already in some of our templates.
His project, ComfyUI-ZFRNodes, folds most of that into two nodes. The Simple Image Generator puts the whole text-to-image and reference-i2i pipeline in one node. The Story Frame Generator turns plain text into per-frame JSON automatically. It runs locally, with optional Ollama for prompts.
What I like is he didn't add features, he just tucked the messy parts everyone keeps rebuilding behind something cleaner.
github.com/zfrsgtcu/ComfyUI-…
Most upscalers split your image into tiles and sharpen each one on its own, with no idea what's around it. It looks great up close but falls apart at full size. The fix is a project called @ComfyUI-42lux-Hildegard-Refiner, built for FLUX.2 Klein.
Comfy Cloud now runs on hundreds of RTX Pro 6000s. Daily active users grew close to 10x after Free Tier shipped. The work spans the core engine, the cloud product, the canvas, and the agent layer being built on top.