The newly established Open Model Initiative, a developer alliance in the field of open-source artificial intelligence, is ambitiously planning to launch a new AI model that may directly compete with Stable Diffusion and its controversial latest version.
Last week, the Generative AI platform Invoke, focused on ComfyUI development at Comfy.org, and the world’s largest stable diffusion repository, CivitAI, launched the Open Model Initiative (OMI). The organization is now leading a community-driven effort to develop open-source AI models for image, video, and audio generation.
Kent Keirsey, CEO of Invoke AI, told Decrypt, “I hope OMI will introduce a new next-generation model, and the community can come together to build support for it in the next three to six months – this is an ambitious goal for a new organization, but I believe we can achieve it.”
OMI aims to develop open-source models of equal or higher quality compared to proprietary models like Ideogram or MidJourney, but without the restrictions of proprietary licenses.
In a recent post, CivitAI claimed, “Comfy tends to utilize a better-designed SD3.”
This move comes shortly after the launch of Stability AI’s SD3, which has been criticized for its restrictive licensing terms. While the stable diffusion series of AI models has been very popular, the SD3 license is considered overly restrictive, leading to its banning by CivitAI.
To prevent this from happening again, the organization advocates for true open development and may adopt licenses such as MIT or Apache-2 to ensure that truly open, unrestricted models meet minimum conditions.
Members stated in an open letter, “We believe that open source is the best way to ensure that AI benefits everyone. Through collaboration, we can offer high-quality, competitive models with open licenses, drive creativity in AI forward, make them free to use, and meet the needs of the community.”
To address ethical concerns, the initiative is also committed to developing a base model without pre-trained capabilities, such as “recognizing unauthorized artist names” and “generating unauthorized personal portraits.”
The initiative has received a lot of support, with over 1000 members joining their Discord server. Keirsey told Decrypt, “We have received over 100 requests for joining and support in just the past 24 hours.”
Regarding funding, Keirsey stated that the initiative will not seek venture capital and will rely on community support and the business model of founding members.
“OMI will not accept investment, because the purpose of this initiative is to establish open models, not to generate profit,” he said. “We have seen what happens when profit becomes the motivating factor behind organizations based on open access principles.”
Each member of OMI will maintain their own business structure, although they are involved in the group.
Robin Ken of ComfyUI told Decrypt that their role in the “democratization of AI” mission is at the tool level, unlike other initiatives such as Stability, Mistral, or Meta, which are more focused on model development. He also confirmed that Comfy is not just an ancillary project and has financial backing.
“Comfy will always remain open-source,” Ken told Decrypt. “We have venture capital support and plan to make money from consulting/business support.”
Ken, Alex Goodwin, and the co-founder of Comfy.org, known as Comfy Anonymous, left Stability AI and are now focusing on the development and growth of Comfy.org along with other developers.
The Open Model Initiative stated that it will initially focus on organizing volunteers, determining governance structures, and managing datasets with community assistance.
LAION, an AI company that compiles image and caption datasets scraped from the web for training several stable diffusions, was initially announced as a founding member of OMI, but was later removed.
Edited by Ryan Ozawa.