VibeVoice 7B (Qwen2.5-7B) — CoreML INT8, fused LM+head, fused diffusion loop, DPM-Solver++ 10-step. Multi-speaker TTS with voice cloning.
Add
vibevoice-coreml to your Swift package. Models auto-download from this repo on first use.
1import VibeVoiceCoreML
2
3let tts = try await MultispeakerTTS(architecture: .model7B)
4let voices = try await tts.encodeVoices(from: [referenceAudioURL])
5for try await frame in tts.speak("Hello world", config: MultispeakerConfig(), voices: voices) {
6 // frame.samples: [Float] at 24kHz
7}
See the
GitHub repo for CLI usage, Python pipelines, and conversion scripts.
