Core ML conversion of stabilityai/stable-audio-open-small for on-device inference on iPhone, iPad and Mac. Converted with coremltools; the packages are stateless, so all sequencing and buffering lives in your Swift code.


Task	text to audio
Upstream	stabilityai/stable-audio-open-small
Packages	4
Download size	1.41 GB
Minimum iOS	17.0
Peak RAM	~1200 MB

Files

File	Size	Compute units	SHA-256
`StableAudioT5Encoder.mlpackage.zip`	94 MB	`cpuAndGPU`	`319a8ba775d30924…`
`StableAudioNumberEmbedder.mlpackage.zip`	367 KB	`cpuAndGPU`	`04bdc5de00a2cf1c…`
`StableAudioDiT.mlpackage.zip`	1.18 GB	`cpuOnly`	`b17da4fc4df85782…`
`StableAudioVAEDecoder.mlpackage.zip`	138 MB	`cpuAndGPU`	`7207544cca9799cc…`
`t5_vocab.json`	732 KB	`-`	`7c9ff3ac1b3dbcaa…`
Total	1.41 GB

compute_units is not a suggestion -- it is the configuration the conversion was verified against. Moving a package to a different compute unit can silently change the numerics (FP16 attention overflow) or crash on the GPU.

Download

bash

hf download mlboydaisuke/coreml-zoo --include "stableaudio/*" --local-dir ./stable_audio
unzip './stable_audio/stableaudio/*.zip' -d ./stable_audio

Use in Swift

swift

import CoreML

let config = MLModelConfiguration()
config.computeUnits = .cpuAndGPU   // as converted — see the table above

// Unzip the .mlpackage, drop it into your Xcode target and Xcode compiles it
// at build time:
let model = try StableAudioT5Encoder(configuration: config)

// ...or compile a downloaded .mlpackage at runtime:
let compiled = try await MLModel.compileModel(at: mlpackageURL)
let model = try MLModel(contentsOf: compiled, configuration: config)

This model is split into 4 Core ML packages that are driven in sequence from Swift. Load them one at a time, copy the outputs out of the MLMultiArray buffers and release each model before loading the next — two large Core ML models resident at once will OOM on an iPhone.

Demo

Sample app — sample_apps/StableAudioDemo, a standalone SwiftUI project.
Models Zoo — this model is downloadable and runnable inside the Models Zoo app on the App Store, no build required.

Conversion

Script: convert_stable_audio.py
Pitfalls hit during conversion (FP16 overflow, ANE buffer limits, stride handling): docs/coreml_conversion_notes.md
Model index: CoreML-Models

License

The conversion inherits the upstream license: Stability AI Community License. See https://huggingface.co/stabilityai/stable-audio-open-small.

Free for non-commercial and limited commercial use; see the Stability AI Community License.

Credits

Upstream authors: stabilityai/stable-audio-open-small, 2024
Core ML conversion: john-rocky (Daisuke Majima)

More models in this format: Core ML Model Zoo — 46 models, each with the recipe that produced it.

Want a different model on-device? Open a request — free, open weights only; the export and its measured numbers get published publicly.

mlboydaisuke/Stable-Audio-Open-Small-CoreML

Stable Audio Open — Core ML