---
title: "init(mlModel:)"
framework: soundanalysis
role: symbol
role_heading: Initializer
path: "soundanalysis/snclassifysoundrequest/init(mlmodel:)-6rmee"
---

# init(mlModel:)

Creates a request that uses a custom sound classification model.

## Declaration

```swift
init(mlModel: MLModel) throws
```

## Parameters

- `mlModel`: A Core ML sound classification model.

## Mentioned in

Classifying Sounds in an Audio File

## Discussion

Discussion The model you provide must accept audio data as input and produce a classification dictionary output that contains the probability of each category. For example, you can generate a sound classifier model by creating an MLSoundClassifier and training it with your own audio files.

## See Also

### Creating a Request

- [init(classifierIdentifier:)](soundanalysis/snclassifysoundrequest/init(classifieridentifier:).md)
- [SNClassifierIdentifier](soundanalysis/snclassifieridentifier.md)