ComfyUI's valuation reached $500 million due to demand from professional creators for precise control over AI generation.

The ComfyUI startup, which provides creators with tools to control the generation of images, video, and audio using diffusion models based on node interfaces, has raised thirty million dollars in a new funding round. As a result of this deal, the company's total valuation has reached half a billion dollars. The leading investor was Craft Ventures, with other venture capital firms including Pace Capital, Chemistry, and TruArrow also participating in the round. According to Marina Temkin's article for TechCrunch AI, this move underscores the industry's growing interest in professional solutions that go beyond simple text prompts to basic models.

The project's history began relatively recently — it originated in 2023 as an open-source initiative, emerging shortly after the widespread adoption of the first diffusion systems. At that time, early versions of popular generative products, including Midjourney and OpenAI's DALL-E algorithms, were barely functional and often made critical errors in object anatomy, such as regularly adding extra fingers to people's hands. To overcome these limitations, the founders developed a modular framework that allowed creators to gain precise control over each stage of the content creation process.

ComfyUI CEO and co-founder Yoland Yan notes that while modern basic models have advanced significantly and no longer add a sixth finger, the need for step-by-step precision has only increased. According to him, typical text-prompt-based solutions only achieve the desired result sixty or eighty percent of the way. Attempting to modify the remaining twenty percent turns into an unpredictable process, which the executive compares to playing a slot machine in a casino. The user is forced to resubmit prompts, which can lead to a completely different final image, including the accidental overwriting of composition elements that were already perfect in the previous generation.

To avoid this unpredictability, ComfyUI offers a node-based workflow. The tool allows visual effects creators to link specific components of the process, ensuring complete control over the quality of the final material. Yoland Yan emphasizes that such a complex architecture of changes cannot be conveyed through a simple text input field in a standard basic model. This approach has resonated widely with creative professionals: the startup currently claims to have over four million users.

The practical significance of the tool is confirmed by its deep penetration into various professional fields. The platform is actively used by experts for creating visual effects, professional animation, developing advertising materials, and in industrial design. The company states that their product has become such an indispensable part of the toolkit for technical artists that specialized positions like "ComfyUI artist" or "ComfyUI engineer" are now frequently found on studio job boards.

Despite the continuous improvement of fundamental models for video and image creation, the startup's management asserts that they are still far from perfect. Yoland Yan is confident that in a world where low-quality content will flood the space, ComfyUI's approach, with mandatory human involvement in the generation cycle, will ultimately capture the most audience attention. The startup will need to develop in a competitive environment, where other players are present, such as Weavy, a company acquired by Figma last year.