---
license: other
license_name: stability-ai-community
license_link: https://huggingface.co/CommuniDev/sd-turbo-coreml/blob/main/LICENSE-stability-ai-community.md
base_model: stabilityai/sd-turbo
pipeline_tag: text-to-image
tags:
  - core-ml
  - sd-turbo
  - text-to-image
  - image-to-image
  - ios
---

# SD-Turbo — Core ML (6-bit, for iPhone)

**Powered by Stability AI.** This is a **converted copy** of [stabilityai/sd-turbo](https://huggingface.co/stabilityai/sd-turbo),
created by **Stability AI**: a fast, distilled version of Stable Diffusion 2.1 that makes a 512 × 512 picture
in 1–4 steps.

CommuniDev LLC did **not** train or fine-tune this model. We only changed its file format so it can run on
an iPhone. It is published so that our iOS apps can download it on first use and run it entirely on the
device: [CommuniDream](https://communidev.com) turns dream journal entries into pictures, and
[CommuniMuse](https://communidev.com) restyles the user's own photos (image-to-image).

## What we changed

Converted with Apple's [ml-stable-diffusion](https://github.com/apple/ml-stable-diffusion) converter
(`torch2coreml`, commit `5a170d2`) from the fp16 safetensors at sd-turbo revision `b261bac6`. No retraining.

- Core ML, attention `SPLIT_EINSUM_V2` (Neural Engine layout), 512 × 512 images.
- UNet and text encoder: **6-bit palettized** weights (k-means). The text encoder's 77 × 77 causal
  attention mask holds `-inf` and is left uncompressed.
- VAE decoder and VAE encoder: fp16. The encoder (added 2026-10-19, from the same conversion run) turns a
  starting picture into latents, for image-to-image.
- Safety checker: the CompVis Stable Diffusion safety checker in Apple's Core ML conversion, unmodified.

Modified files: `TextEncoder.mlmodelc`, `Unet.mlmodelc`, `VAEDecoder.mlmodelc` and `VAEEncoder.mlmodelc`
are converted, modified forms of the original checkpoint. Details are in `NOTICE`.

## Files

| Path | Size |
|---|---|
| `compiled/Unet.mlmodelc` | 623 MB |
| `compiled/TextEncoder.mlmodelc` | 304 MB |
| `compiled/VAEDecoder.mlmodelc` | 95 MB |
| `compiled/VAEEncoder.mlmodelc` | 65 MB |
| `compiled/SafetyChecker.mlmodelc` | 580 MB |
| `compiled/vocab.json`, `compiled/merges.txt` | 1.3 MB |

`SHA256SUMS` lists a SHA-256 hash for every file. Check it with `shasum -a 256 -c SHA256SUMS`.

## How it is used

With Apple's Swift `StableDiffusionPipeline`: DPM-Solver++ (linspace spacing, which starts at t = 999),
2–4 steps, guidance 1.0–2.0, seed-reproducible. At guidance 1.0 the negative prompt has no effect.
For image-to-image, pass a 512 × 512 starting image and a strength; the steps actually run are
steps × strength, so keep that at 1 or more.
**The safety checker must stay on**: CommuniDream and CommuniMuse refuse to run without it.

## Limitations

- 512 × 512 only (the shape is fixed at conversion).
- Distilled for speed: less detail than Stable Diffusion 2.1 at 20+ steps, and weaker style control.
- It can produce wrong, biased or disturbing pictures, and text in pictures is unreliable.
- The safety checker can miss unsafe pictures and can block harmless ones.

## Licence

`TextEncoder`, `Unet`, `VAEDecoder` and `VAEEncoder` are under the **Stability AI Community License**
(`LICENSE-stability-ai-community.md`, word for word): use must follow Stability AI's
[Acceptable Use Policy](https://stability.ai/use-policy), and commercial use requires registration with
Stability AI. `SafetyChecker` is under **CreativeML Open RAIL-M** (`LICENSE-creativeml-openrail-m.txt`).
See `LICENSE.md` and `NOTICE`.

## Citation

```
@article{sauer2023adversarial,
  title   = {Adversarial Diffusion Distillation},
  author  = {Sauer, Axel and Lorenz, Dominik and Blattmann, Andreas and Rombach, Robin},
  journal = {arXiv preprint arXiv:2311.17042},
  year    = {2023}
}
```