ABrush
Generate and edit AI images directly inside Adobe Photoshop, with 23+ built-in models and professional generation controls.
About ABrush
ABrush is an AI image generation tool integrated into Adobe Photoshop. Users can generate images directly inside Photoshop and edit them within their documents, without switching between multiple tools, keeping the workflow focused. It includes 23+ mainstream models (including Stable Diffusion, Flux, Gwen Image, and more) and 5 AI model modes, along with professional controls such as ControlNet, IP-Adapter/reference images, generation history, presets and preset sharing, and LoRA styles. According to the official website, users own every image they generate, user data is not used for training, and commercial projects enjoy full privacy.
Best for
Best use cases
Generate multiple image variants for a design project directly in a Photoshop document and filter through them stage by stage
Control generation results using ControlNet, reference images, and LoRA styles
Embed AI generation into an existing Photoshop workflow to avoid switching back and forth between tools
Features
In-Photoshop generation
Generate multiple variants directly in Photoshop and choose a different model for each stage of your workflow, without switching tools.
Multi-model support
Includes 23+ mainstream models and 5 AI model modes, covering different generation needs.
Professional generation controls
Offers controls such as ControlNet support, IP-Adapter/reference images, generation history, presets and preset sharing, and LoRA styles.
Multiple interface modes
Offers 3 interface modes, integrated with artists' workspaces.
Pros
- Generate and edit images directly inside Photoshop without switching tools
- Includes 23+ built-in models, letting you pick the right model for each stage of your workflow
- Provides professional generation controls such as ControlNet, IP-Adapter/reference images, and LoRA styles
- Supports presets and preset sharing, and keeps generation history
- Per the official website, users own the images they generate and user data is not used for training