Packaging Expert Workflows for Creators
Coco Mao, CEO of OpenArt, on building the TikTok for AI video
This points to a fork in generative media product strategy, between building for experts who want to wire models together themselves, and building for mainstream creators who want one button that just works. A workflow marketplace could have turned OpenArt into the bridge between those groups, with experts packaging multi step image and video pipelines for newer users, but it also would have pulled the product toward the complexity of ComfyUI rather than toward a simpler creator app.
-
ComfyUI already has the raw ingredients for this model. It treats workflows as node graphs, supports built in and community templates, and lets custom node authors ship example workflows inside the product. That makes reusable workflows possible, but still leaves users dealing with graphs, models, and dependencies.
-
Other workflow companies show what the more polished version looks like. n8n has a template library and is building a creator marketplace, while Segmind turns visual AI workflows into pre built industry pipelines and even deployable APIs. The pattern is expert built logic, wrapped in a friendlier surface for everyone else.
-
OpenArt chose the opposite path. The product already packages image editing into pre baked flows like sketch to image, upscaling, and face replacement, and it monetizes with simple credit subscriptions. In practice that means hiding the graph and selling outcomes, which fits hobbyists, creators, and SMBs better than selling workflow construction.
The next wave in AI video will likely reward companies that absorb expert workflows into simple products rather than exposing the workflow layer directly. OpenArt is positioned to keep turning messy multi model pipelines into creator friendly features, while the workflow ecosystem itself is more likely to remain infrastructure for power users, developers, and enterprise tooling.