whisperkit-coreml

OpenAI Whisper in CoreML format, converted and published by Diction Labs for on-device speech-to-text on Apple hardware.

Use with WhisperKit

let config = WhisperKitConfig(
    model: "openai_whisper-base",
    modelRepo: "DictionLabs/whisperkit-coreml"
)
let pipe = try await WhisperKit(config)

Runs on argmax-oss-swift (MIT). Tokenizers are resolved separately from the matching openai/whisper-* repo, so a first run still needs network access even with the weights already on disk.

Variants

Folder	Base model
openai_whisper-base	openai/whisper-base
openai_whisper-small	openai/whisper-small

Provenance

Converted from OpenAI's PyTorch checkpoints with whisperkittools (MIT), not repackaged from another CoreML distribution. Published under our own namespace so the models Diction downloads come from a repository we control: a new upstream release reaches nobody until we have tested it and promoted it deliberately.

MIT licensed, as is Whisper itself.