How to transcribe a Zoom call on Mac in 2026 (and what's new this year)
If you want a usable transcript of a Zoom, Meet or Teams call on your Mac without inviting a bot, you have five real options. Here is when to use each one, and how Yapper's Both Sides capture changed the calculus this year.
TL;DR: There are five real ways to transcribe a Zoom, Google Meet, Microsoft Teams or FaceTime call on macOS in 2026. The right one depends on whether you’re willing to (a) join a bot to the meeting, (b) record system audio, or (c) post-process a recording. Yapper’s Both Sides mode — new in 2026 — captures system audio and your microphone in parallel without ever joining the call. We’ll go through all five.
Why this got harder before it got easier
Apple tightened the screen-recording entitlement in macOS Sonoma, which broke the older trick of routing system audio through a virtual loopback like Soundflower or BlackHole and pointing a transcription app at it. Most users who tried that workflow in 2024 gave up.
The 2026 picture is cleaner. ScreenCaptureKit now exposes audio-only capture with a single user permission grant, which means a dictation app can listen to whatever’s playing through your speakers — Zoom, Meet, Teams, FaceTime, a Twitter video, a podcast, anything — without you having to install kernel extensions. Yapper uses this. Most older transcription tools do not.
The five options
1. Zoom’s built-in transcription
Free for paid Zoom accounts. The host has to enable “Cloud recording” with transcription. Output: a text file emailed after the call.
Use it when: you’re hosting and don’t need real-time text. Don’t use it when:you’re a guest, the host hasn’t enabled it, or you’re on Meet, Teams or FaceTime.
2. Otter.ai or similar bot-in-the-meeting tools
Otter, Fireflies, and Read all work the same way: their bot joins your meeting as a participant, records the audio, and produces a transcript. It works on every major platform.
Use it when: you want a hands-off transcript and don’t mind a third party joining. Don’t use it when: the meeting is sensitive and inviting an AI bot is awkward, or your client/legal team doesn’t allow it.
3. Krisp / Granola style local transcribers
These run on your Mac, capture system audio + mic, and produce a transcript locally or via the cloud. Granola in particular has captured a lot of the executive market in 2025-2026.
Use it when: you want a polished post-meeting summary and don’t care about realtime.Don’t use it when: you also want fast single-shot dictation in the same app, or want everything to live in one searchable timeline alongside your other voice work.
4. QuickTime + manual transcription
Free, built-in. Record your screen with audio in QuickTime, then feed the file to Whisper, MacWhisper or any cloud ASR. Quality is fine. Latency is hours.
Use it when: you only need a transcript occasionally and don’t want a subscription.Don’t use it when: you need it routinely.
5. Yapper Both Sides (Max plan)
Yapper’s Both Sides mode captures system audio through ScreenCaptureKit alongside your microphone, in parallel, on-device. It transcribes both streams, separates speakers (you vs them), and saves the whole conversation to your timeline searchable with full text.
Use it when: you want a real-time transcript of any call — Zoom, Meet, Teams, FaceTime, a phone call routed through your Mac, even a video on Twitter — without inviting a bot, and you want all of it indexed alongside your normal dictation.
Don’t use it when: you don’t want a searchable transcript at all, or you’re on Free or Pro tier (Both Sides is Max-only). See pricing.
Comparison at a glance
| Tool | Realtime | No bot | Works on Meet/Teams/FaceTime | Price |
|---|---|---|---|---|
| Zoom built-in | Yes | Yes | No | Free w/ paid Zoom |
| Otter / Fireflies | Yes | No (bot) | Yes | $10-30/mo |
| Granola / Krisp | Mostly | Yes | Yes | $10-25/mo |
| QuickTime + Whisper | No | Yes | Yes | Free |
| Yapper Both Sides | Yes | Yes | Yes | $29/mo Max |
Step-by-step: Both Sides on Yapper
- Install Yapper for macOS and upgrade to Max plan.
- On first launch, grant microphone access (standard) and screen recording access. Yapper only uses screen recording for audio capture — the video stream is discarded.
- Open the meeting in any app: Zoom, Meet, Teams, FaceTime, or even a video on a website. No bot needed.
- Press ⌥+Z to start a Both Sides session. A small indicator appears in the menu bar.
- Talk normally. Yapper transcribes both you and the other side in real time. Speaker labels are inferred from microphone vs system audio.
- Press ⌥+Z again to stop. The session lands in your timeline, fully searchable, with timestamps and speaker labels.
Privacy
Both Sides is local-first. The audio is transcribed on your device using Whisper-large-v3-turbo. The raw audio never leaves your Mac unless you explicitly opt into encrypted cloud backup (Max plan, off by default).
Some jurisdictions require both-party consent before recording. The legality is on you — we just provide the tool. Where consent is required, Yapper’s indicator is visible enough that polite practice is to mention it at the top of the call.
Frequently asked questions
Does Both Sides work on Google Meet?
Yes. Both Sides captures whatever audio your Mac is playing regardless of source — Meet, Zoom, Teams, FaceTime, Slack Huddles, Discord, or browser-based platforms.
Will the other side know I’m transcribing?
Not unless you tell them. Both Sides runs locally and doesn’t join the call. As above — check the laws in your jurisdiction and ideally just mention it.
Can I transcribe a phone call on my Mac?
Yes, if the call is routed through your Mac — FaceTime audio, a SIP softphone, Continuity calls from your iPhone. Standard PSTN calls on the iPhone alone are not yet supported (iOS-side capture is a different story; see the iOS rollout note on pricing).
How accurate is Both Sides on noisy calls?
Whisper-large-v3-turbo handles compressed VoIP audio well — we see word-error-rates of 6-9% on typical Zoom audio, dropping to 3-4% with a custom dictionary loaded for names and jargon. See our broader latency benchmarkfor context on accuracy vs speed trade-offs.
Want to try the fastest dictation tool we measured? Download Yapper for macOS — 2,500 free words, no card. Or read the next post: /blog.