Download
Menu

Product

FeaturesCaptions, dictation, meetings, translation, notes, sharing, and accessibility tools. PricingFree, Pro, Team Packs, and education pricing. TeamsTeam Packs for global calls and shared activations.

Use Cases

Use CasesPrivate captions, dictation, translation, and transcripts for real work. Zoom Captions on MacCaption Zoom without host settings or meeting bots. Hearing AccessibilityPrivate captions for Mac audio, calls, and lectures. Legal TranscriptionLocal-first captions and transcripts for legal work. Research InterviewsFollow interviews and keep local transcripts.

Resources

GuideSet up captions, translation, audio modes, and exports. Compatible AppsSee which Mac apps and audio sources NotchLive can caption. BlogGuides for live captions, translation, and private transcription. Best Live Caption AppsCompare private and cloud caption tools. Release NotesSee what changed recently. AboutLearn who makes NotchLive and why it exists.

Kōkako

Rebuilt for calls. Kōkako 2.0 is designed to reduce matching Mac playback in NotchLive’s microphone path without taking over the processing your meeting app applies to your voice.

Download NotchLive Free

Your meeting app should own your microphone. NotchLive should listen.

Kōkako 1.0 used macOS Voice Processing, which could compete with processing in Zoom, Meet or Teams. Kōkako 2.0 separates those jobs.

Separate capture, cancellation and recognition.

Kōkako 2.0 gives each part of the call audio path one clear job.

Microphone captions with Kōkako need Screen Recording permission for the local playback reference. If the reference or AEC3 cannot start, a microphone-only caption session uses Legacy without changing the saved engine. Everything mode still requires system-audio capture. Voice Notes use Legacy in this release and do not run through AEC3.

What AEC3 does in isolation.

A 12-second synthetic test at 16 kHz measured speaker-echo rejection and voice preservation before speech recognition. It is a DSP test, not a real meeting benchmark.

Built for calls, honest about the boundary.

Kōkako owns echo cancellation inside NotchLive. Your meeting app owns the voice processing heard by everyone else.

Kōkako processing stays on your Mac.

Within NotchLive, the microphone signal, ScreenCaptureKit playback reference, AEC3 processing and WhisperKit recognition stay on your Mac. NotchLive never includes audio, captions, transcripts, translations, notes or meeting content in diagnostics.

Questions about the Kōkako voice engine.

What changed in Kōkako 2.0?

Kōkako 2.0 replaces macOS Voice Processing with a normal microphone path and local WebRTC AEC3 echo cancellation inside NotchLive.

Does Kōkako change how people hear me in a meeting?

Kōkako 2.0 does not enable Apple VoiceProcessingIO. Your meeting app keeps control of the microphone processing it sends to other participants.

Does Kōkako remove background noise?

No. Kōkako 2.0 performs echo cancellation only. It does not add noise suppression or automatic gain control, so room noise can still affect microphone captions.

Why does microphone captioning need Screen Recording permission?

ScreenCaptureKit provides a local reference for audio playing from your Mac. AEC3 uses that reference to help reduce matching playback in microphone captions. The reference stays on your Mac.

What happens if AEC3 cannot start?

For a microphone-only caption session, NotchLive uses Legacy without changing the saved engine. Everything mode still requires system-audio capture. Legacy also remains selectable under Settings, then Audio.

Do Voice Notes use Kōkako 2.0?

Not currently. Voice Notes use the Legacy microphone path and do not run through AEC3.

Which macOS versions support Kōkako 2.0?

Kōkako 2.0 supports macOS 14 and later on Apple silicon and Intel Macs. Live translation and Meeting Intelligence have their own separate system requirements.