Waki Recorder features
01 / Record
Both sides of the call, in the background.
Start from the window or the menu bar. Waki Recorder captures your microphone and the system audio together, watches the level so you can see it is working, and writes the recording in chunks so nothing is lost if the call runs long.
A meeting, from start to minutes
- Microphone and system audioBoth are recorded, so the other participants are in the transcript too.
- Live level and silence detectionThe window shows the audio level and warns when nothing is coming in.
- Upload a file insteadDrop an existing recording in and it goes through the same transcript-and-minutes pipeline.
02 / Transcript & speakers
A transcript as it happens. Speakers named afterwards.
Transcription runs on your Mac while you record, in thirty-second chunks, so the transcript is nearly complete when you stop. Speaker identification runs on the full audio afterwards; rename “Speaker 1” to a person once and every line and the minutes follow.
From audio to a readable transcript
- Streaming transcriptionChunks are transcribed as they finish, with overlaps de-duplicated, so the final pass is short.
- Speaker diarizationSpeakers are separated on the full recording and mapped onto the transcript.
- Rename speakers onceChange a speaker’s name and it is replaced throughout the transcript and the minutes.
03 / Minutes & export
Minutes you can send, in the tools you already use.
The transcript goes to the model you choose, through Waki AI Service or your own Claude, OpenAI or Ollama, and comes back as minutes with a generated title. Export them, or push them straight to your notes.
Where the minutes go
- Your model, your keyUse Waki AI Service or connect Claude, OpenAI or a local Ollama model. Regenerate the minutes from the same transcript any time.
- Export and sharePDF, HTML or Markdown, or push to OneNote, Notion or Google Drive from the meeting page.
- A meeting librarySearch, pin and tag meetings; each one keeps its audio, transcript, minutes and processing log.