Changelog

What's new in Speaklone.

v1.1

March 2026

• •System-wide speech recognition (ASR) with live transcription overlay
• •Dictation works across any app via macOS Services and keyboard shortcuts
• •Local HTTP API server and URL scheme for external TTS integration
• •Floating speaking overlay for system-event TTS
• •Timbre lock — save and reuse the exact voice characteristics of a designed voice
• •Incremental ASR encoder caching for faster recognition
• •Voice designer coherence improvements for longer passages
• •Auto-downloads essential voice model on first launch

v1.0.1

February 2026

• •Fixed crash with longer voice reference clips
• •Hotkeys now work with all voice types
• •Improved model discovery and download reliability on iOS

v1.0

February 2026

• •Initial release
• •Voice cloning from 3+ seconds of audio
• •Voice design from text descriptions (Mac)
• •9 expressive preset voices
• •10 languages
• •Runs entirely on-device via MLX on Apple Silicon