110 (20260425)

Transcription & Audio Analysis

• Sequence now classifies audio regions to distinguish speech from silence and other sounds before transcription begins, which improves accuracy dramatically.
• CoreML model handling is more robust with better runtime detection reducing edge-case failures.
• Post-processing is significantly improved to handle natural speech patterns better.

Search

• Your selection is even more reliable.

History

• Timing breakdowns now show three distinct phases, giving you clearer insight into where time is spent during processing.

Export to FCP

• Crossfade duration handling is more precise when removing filler words, avoiding occasional timing glitches in Final Cut Pro timelines.

Accessibility

• More UI elements now carry accessibility identifiers, improving VoiceOver support and automated testing coverage.

Speed & Stability

• Variant switching is now even smoother.
• General cleanup of edge cases and internal consistency