Install draw-things with Homebrew Cask

Project links

• Homepage
• Package manager page

Commands and aliases

No executable aliases were found in the local package database.

Project history

Draw Things began as an Apple-platform application centered on private, local diffusion-model inference. Official documentation emphasizes local processing and an interface for prompts, models, LoRAs, ControlNet, textual inversion, Core ML, and scripts.

The project later published substantial public-facing inference, sampler, data-model, and training code in drawthingsai/draw-things-community while retaining its primary application in a private monorepository.

Adoption history

Draw Things developed a community ecosystem around downloadable models, LoRAs, scripts, documentation, and public inference components. Its official public repository welcomes contributions while explaining that the application itself is managed in a private monorepository and selected public-facing code is synchronized outward.

The application expanded from local image generation on Apple devices to training, model import, scripting, video and audio-capable workflows, optional cloud compute, server offload, and a command-line interface.

How it is used

Users download or import a supported model, enter positive and negative prompts, select generation parameters, and generate or edit media locally. Advanced workflows use LoRAs, ControlNet, masks, training, JavaScript automation, server offload, or the CLI.

Local inference protects privacy and avoids a mandatory remote GPU service, while optional Draw Things+ and cloud-compute features cover workloads that exceed local hardware capacity.

Why package nerds care

Draw Things packages a rapidly changing generative-media stack as a native Apple application, avoiding much of the Python-environment and GPU-toolchain setup associated with Stable Diffusion front ends.

Its public community repository, Homebrew tap, CLI, model catalogs, scripting API, and self-hosted gRPC server make it unusually interesting at the boundary between consumer Mac casks and hackable local-AI tooling.

Timeline

• 2023: Official documentation covered local AI image generation, models, LoRAs, ControlNet, Core ML, and scripting.
• 2025: Draw Things introduced optional cloud-compute workflows alongside its established local generation model.
• 2026: Releases added newer image and video model families, Apple Neural Engine optimizations, Lightning Draft, and expanded training and editing support.

Related projects

• Stable Diffusion and other supported generative-media model families provide the models Draw Things executes.
• The official ecosystem includes draw-things-community, community-models, community-scripts, community-docs, media-generation-kit, and a ComfyUI gRPC extension.

Sources

• https://docs.drawthings.ai/
• https://drawthings.ai/downloads/
• https://github.com/drawthingsai
• https://github.com/drawthingsai/draw-things-community