Ant Group's inclusionAI open-sourced Ming-Image-0.1-Design-Layer for RGBA layer decomposition
inclusionAI (Ant Group) released Ming-Image-0.1-Design-Layer, an image-text-to-image model that decomposes a flattened design image into a requested number of RGBA layers, under the MIT License on Hugging Face. The source release was published on 2026-09-17.
China context
- Original name
- 蚂蚁集团
- Outside China
- Open weights · huggingface.co
- Claims
- Company-reported; not yet independently evaluated
- For builders
- Developers outside China can download the model weights from Hugging Face and integrate layer decomposition into their own applications under the MIT License.
- For investors
- Ant Group's release of open-weights design models signals a strategic move into the creative tools market, potentially challenging established design software vendors.
inclusionAI (Ant Group) released Ming-Image-0.1-Design-Layer on Hugging Face under the MIT License. The model takes an image and a layer plan to decompose a flattened design into a requested number of RGBA layers. It is part of the Ming-Image series, which also includes Ming-Image-0.1-Design, a 6B text-to-image model for UI, infographics, posters, and other text-rich visual designs with RGBA output support.
Ming-Image-0.1-Design-Layer uses a working-resolution bucket of 1024 (or 512 for faster decomposition), 12 sampling steps, CFG scale 2.0, BF16 precision, and requires one CUDA GPU with 80 GiB VRAM. The model outputs standalone RGBA PNG files. Prompt enhancement can use Ling-3.0-flash-VL or qwen3.8-27B. The companion Ming-Image-0.1-Design model recommends 2048x2048 resolution, 12 sampling steps, CFG scale 1.0, BF16, and the same hardware.
Designers and developers outside China can now use an open-weights model to automatically separate flattened design images into editable RGBA layers, reducing manual layer extraction work. This directly competes with proprietary design tools that offer layer decomposition as a paid feature.
The MIT License allows commercial use, modification, and redistribution, enabling businesses to integrate layer decomposition into their design pipelines without licensing fees. The model's focus on UI, infographics, and posters targets the growing market for automated design generation.
Next signals to check: whether the model is integrated into popular design tools or workflows, and whether the Crello test set benchmark results are independently reproduced. Also monitor if inclusionAI releases larger or more specialized design models.