About Us
We are a team dedicated to iterate and improve ComfyUI, support the ComfyUI ecosystem with tools like node manager, node registry, cli, automated testing, and public documentation.
The Vision
Open source AI model will win over the long run against closed models and we are at the beginning of it. Yet if open source models can only be utilized by a handful of cloud providers or closed sourced tools, is AI truly democratized? Our core mission is to advance and democratize AI tooling. We believe that the future of AI tooling is open-source and community-driven.
The Team
Contact Us
Frequently Asked Questions
Comfy will continue to adopt state of the art models to stay on the bleeding edge. We also aim to review PRs and address Github issues faster. There is also a need to improve the user experience and developer experience for custom node authors. While we are still iterating on the exact model, we plan share future improvements under Projects on Github. All major changes will begin as a Github Discussion, which will be shared on our Discord and Matrix servers. Once we reach a final design we will add the implementation to a public roadmap. Please follow our twitter, discord or matrix servers for updates.
We currently do not make any money. We are backed by supporters and investors who are aligned with our vision for developing open-source AI tooling, if you are one, please reach out to us at hello@comfy.org. We do intend to make money in the future.
We don't yet. We plan to make money through consulting and enterprise support. It's a proven model (e.g. RedHat). If you can build amazing OSS software that's widely adopted as the industry standard, you can finance the core development while empowering an entire ecosystem. We will never hide/close source features behind a paywall or license.
Nodes on the registry are semantically versioned and scanned for malicious behavior. While we can't catch everything, we will soon add checks to see if nodes include custom pip wheels, execute arbitrary sys calls or call outside APIs. A human will review flagged nodes. More importantly, when a security concern is found we will ban the node and notify any users who have installed it. This process will be improved over time.
ComfyUI is an application, a backend, and a developer platform. We invest in tools to make using ComfyUI safe and reliable. We host Comfy Registry, the repository hosting custom nodes. Nodes on the registry are semantically versioned and scanned for malicious behavior. While we can't catch everything, we will soon add checks to see if nodes include custom pip wheels, execute arbitrary sys calls or call outside APIs. We also host a Continuous Integration Testing Suite that tests new commits in Comfy against workflows on various operating systems and GPUs to make Comfy more reliable to use.
1. Transparency and communication. 2. Open-source and community-driven. 3. Democratizing AI.
Despite the recent turmoil in OSS AI commmunity, there are huge amount of advancement in OSS AI models. We are working closely with OSS model builders to bring the best models to ComfyUI. We also plan to put more resources into AI models in the future.
You can voice your feedback and participate in our discord/matrix channel. If you can also help submitting a bug report or feature request, please create an issue on Github. Many issues are tagged #good-first-issue , which are good for new contributors to Comfy. Otherwise please submit a PR and we will review it. In the near future, we will also introduce a governance structure similar to other OSS projects.
You can follow us on Twitter, join our Discord or Matrix channel.