Aurora Mobile's Modellix.ai Integrated with ComfyUI

On October 30, 2026, Aurora Mobile (NASDAQ: JG) announced that Modellix.ai — its platform for accessing media generation models — now works inside ComfyUI. The company reported this in a press release. ComfyUI is a visual editor based on graphs (nodes): the user builds a chain of steps with a mouse rather than with code.

Previously, Modellix.ai provided unified access to image, video, and sound models through its own interface or API. Now the same models can be called in a node environment. The connection uses the existing Modellix.ai API key: the user selects a live model, configures its current parameters, attaches media inputs, and sends the result to ComfyUI's standard preview and save nodes.

The difference is that previously, each model provider in ComfyUI usually required a separate package. Here, the developer works with live model schemas (descriptions of inputs and parameters) and standard media connections without building a separate package for each vendor. This is the company's statement; there is no independent verification in the note.

In practice, the integration supports model selection, media upload, ordered reference inputs, asynchronous generation (the request goes into a queue and the result is retrieved later), and retrieval of already created tasks. Included are importable examples: image generation, frame-to-video transition, speech synthesis, and task recovery. A creator can build steps for image, video, and sound on a single canvas and reuse that schema in other projects.

The press release includes a quote from Modellix.ai founder and CEO Chris Lo saying that the further spread of generative media will depend on how well the technology integrates into workflows, no less than on model quality. The full quote in the note is cut off.

What remains unknown: which specific models and providers are available through the integration, how much it costs, and how the limits are structured. There is no data on this in the note.