Engineering @ Draw Things

2 Days to Ship: Codex-authored Metal Compute Shaders in Draw Things

In the 1.20260314.0 release, we shipped our first two Codex-authored Metal compute shaders, which improved LTX-2.3 video VAE decoding speed by about…

19 hrs ago • Authors of Draw Things

3

1

September 2025

Optimizing Qwen Image for edge devices

Qwen Image is the largest open-weight state-of-the-art image generation model. With its 20B parameters, it poses unique challenges for Draw Things to…

Sep 5, 2025 • Authors of Draw Things

2

2

April 2025

BF16 and image generation models

Draw Things maintains its own local inference and training stack for image generation models.

Apr 30, 2025 • Authors of Draw Things

2

January 2025

Metal FlashAttention 2.0: pushing forward on-device inference & training on Apple silicon

Metal FlashAttention underpins Draw Things’ claim of fastest image generation inside the Apple ecosystem.

Jan 7, 2025 • Authors of Draw Things

June 2024

From iPhone, iPad to Mac - enabling rapid local deployment of SD3 Medium with s4nnc

SD3 Medium was released on June 12th, 2024.

Jun 24, 2024 • Authors of Draw Things

October 2023

Draw Things democratizes local large model fine-tuning on iPhone, iPad and Mac

Large language models and image generation models currently require hundreds of thousands to tens of millions of dollars for hardware acquisition and…

Oct 5, 2023 • Authors of Draw Things

1

August 2023

Integrating Metal FlashAttention: accelerating the heart of image generation in the Apple ecosystem

Draw Things was the first practical app to run full-blown image generation models at the “edge” — directly on your mobile phone.

Aug 9, 2023 • Authors of Draw Things