Knows when a meeting starts
Confidence-scored detection across Zoom, Teams, Meet and FaceTime: microphone use, calendar, window structure. Or press ⌘R.
Cadenza records the call, separates the speakers, and writes the summary, decisions and action items. On-device by default. Bring your own AI keys when you want a cloud model, and pick the work up again from any browser.
From the moment the call starts to the recap at the end of the week.
Confidence-scored detection across Zoom, Teams, Meet and FaceTime: microphone use, calendar, window structure. Or press ⌘R.
Apple Speech and Whisper on-device. OpenAI and Gemini when you want a cloud model, including live captions while you are still talking.
Name a voice once and Cadenza attributes their lines in later meetings. Speaker-time distribution on every transcript.
Structured output with assignee, deadline and priority. Claude, GPT, Gemini, MiniMax or Apple Foundation Models. Your prompt, if you want one.
The assistant picks the time range, speakers and keywords from your question, then fills a token budget with the right excerpts. Weekly and monthly recaps too.
Notion, Craft and Markdown, one click or automatic. An MCP server hands transcripts to Claude Code, Codex and friends, off by default.
Web Sync is optional and available in private beta. Turn it on and your library follows you. With a provider key in the vault, the web app can re-transcribe, re-summarize, translate, answer questions and build recaps, using the same providers you set up on the Mac.
Free while in beta. No account required for the Mac app. Web Sync requires Google sign-in when enabled and uses Cadenza-operated storage.