models/Shuttle 3 Diffusion - bf16

Shuttle 3 Diffusion - bf16

7/26/2025

5:13:25 AM

Related Keywords & Tags

bf16,checkpoint,complex prompts,flux.1,flux.1 s,image quality,photorealistic,resource efficiency,shuttle 3 diffusion,shuttleai,style,text-to-image ai model,typography

A colossal golden clock floats above a city of tall spires and towers, surrounded by birds and glowing rays of light shining through clouds.

A solitary figure standing on rocky terrain surrounded by misty landscape and mysterious jagged mountains under a twilight sky filled with dense, ethereal clouds.

A digitally rendered portrait of a woman wearing a shiny black top hat, holding sheet music in front of a historical urban 1800s street scene with horse-drawn phaetons.

A towering ethereal robotic being with galactic head and glowing tendrils stands in a crumbling futuristic city, while a figure in an exosuit looks up at it beneath a neon sky.

A cybernetic woman with intricate glowing circuitry and pixel decorations on her face, surrounded by robotic figures with glowing eyes in a pixelated, futuristic cityscape.

A fluffy black cat with bright yellow eyes sitting on a wooden swing in the rain during a stormy evening, surrounded by dark clouds and lush greenery.

Tips

Model enters a refiner mode beyond 10 steps, improving image details without changing composition.

Creator Sponsors

You can try out the model through a website at https://chat.shuttleai.com/images

Use Shuttle 3 Diffusion via API through ShuttleAI with Documentation

Shuttle 3 Diffusion

Shuttle 3 Diffusion is a text-to-image AI model designed to create detailed and diverse images from textual prompts. It offers enhanced performance in image quality, typography, understanding complex prompts, and resource efficiency.

You can try out the model through a website at https://chat.shuttleai.com/images

Using the model via API

You can use Shuttle 3 Diffusion via API through ShuttleAI

- ShuttleAI

- Documentation

Shuttle 3 Diffusion uses Flux.1 Schnell as its base. It can produce images similar to Flux Pro in just 4 steps, and it is licensed under Apache 2. The model was partially de-distilled during training. When used beyond 10 steps, it enters "refiner mode," enhancing image details without altering the composition. We overcame the limitations of the Schnell-series models by employing a special training method, resulting in improved details and colors.