Webex Transcription on Mac: The On-Device, No-Bot Way

Webex is the meeting app a lot of people use because their employer or their customer told them to. Cisco still owns large pockets of healthcare, federal contracting, finance, legal, and any vertical where IT bought a Cisco contract a decade ago and never unwound it. If that’s the world you work in, you live in Webex whether you wanted to or not — and the “can I just get a transcript of this call” question runs into the same wall it does on Zoom or Teams.

Webex does have native transcription. It is also gated behind Webex Assistant, a paid add-on that the host’s Webex plan needs to include, and the host has to turn it on for the meeting. If you are a guest joining a customer’s Webex, or you are on the personal Webex plan, or your employer bought Webex Meetings without the Suite add-on, there is no checkbox you can flip on your end to get a transcript back.

The usual fallback is to invite Otter, Fireflies, or Read.ai’s bot to the call. That works. It also means a third party is logged into your customer’s Webex call and shipping every minute of audio to a cloud transcription pipeline. For some industries Webex is popular in specifically because of data-handling rules, that fallback is the wrong default by construction.

This post is about the third path: capturing a Webex call locally on a Mac running macOS 26 Tahoe, transcribing it on-device, and ending up with a Markdown transcript and an audio-anchored summary that never left the laptop. No Webex Assistant. No bot. No cloud upload.

What Webex gives you natively, and where it stops

Worth being precise about the gap, because Webex’s plan tiers are not intuitive:

Webex Free / Personal. No transcription. 40-minute meeting cap on three or more participants. Recording is local only and saves an MP4.
Webex Meet ($14.50/user/mo). Cloud recording, no Webex Assistant transcription unless you separately add the Suite or AI Assistant SKU.
Webex Suite ($25/user/mo). Includes Webex AI Assistant, which transcribes meetings, generates summaries, and labels speakers from the calendar context.
Webex Enterprise. Negotiated. Usually includes Suite features plus data-residency controls.

A few practical consequences:

If you are a guest on someone else’s Webex (very common in enterprise sales calls, vendor pitches, and customer onboarding), you have no host permissions. You cannot turn on Webex Assistant. You either get the transcript the host’s plan generated and the host is willing to share, or you get nothing.
If you are the host but on Meet without Suite, you can pay an extra $10.50/user/mo for Suite, or you can find another way to get the transcript.
If your IT department disabled Webex AI Assistant for compliance reasons — which is common in healthcare, finance, and federal — the feature is dark even on plans that technically include it.

If you have Webex Suite licensed and IT lets you use it and you are the host on the calls that matter, you do not need any of what follows. The rest of this post is for the case where one of those three conditions doesn’t hold.

The on-device alternative on a Mac

Two macOS frameworks introduced in macOS 26 Tahoe do all the heavy lifting:

ScreenCaptureKit with audio-only content filters. Captures the system audio output of a specific running process — in this case, the Webex Mac app, or whichever browser has the Webex web client open. Audio capture is scoped to that process, so nothing else on the Mac (your Music app, Slack call sounds, an AirPods notification) ends up in the recording.
SpeechAnalyzer, Apple’s on-device automatic speech recognition framework introduced at WWDC 2025. Apple’s published benchmarks have it ~55% faster than Whisper v3 Turbo on the same Apple silicon, running entirely on the Neural Engine, with no network round trip.

These are the same primitives used to record Zoom calls without a bot and to capture Teams meetings on Mac. ScreenCaptureKit doesn’t care which meeting app is producing the audio — it sees Webex, Zoom, Teams, Meet, Discord, and a podcast playing in Safari the same way.

Dictanta is the Mac/iPhone/iPad/Vision Pro app that wires the two frameworks together with Apple’s Foundation Models on-device LLM for the post-meeting summary. The rest of this post walks the Webex-specific flow.

The full Webex capture flow

Assume a fresh install on macOS 26.

1. Join the Webex call however you normally do

Webex runs as a native macOS app, a web client, and a Chrome extension. All three send audio to the Mac’s audio output the same way. You can be signed in to a personal Webex account, an enterprise SSO, or joining as a guest with no Webex account at all. The capture path is the same.

You don’t need Webex Assistant licensed. You don’t need to be the host. You don’t need the meeting organizer to enable transcription. None of those gates apply to capturing the system audio your Mac is already producing.

Dictanta lives in your menu bar. Default global hotkey is ⇧⌘R. Press it. The icon turns coral and pulses. A short live-captions strip drops down under the menu bar with the transcript scrolling in real time. The strip is invisible to a Webex screen share if you have one running, but you can also hide it entirely with ⌘⇧K.

The first time you record system audio, macOS prompts for screen recording permission. This is Apple’s TCC entitlement gating — ScreenCaptureKit’s audio path lives under the same entitlement as the video path, even though no video frames are captured. Grant it once, never again.

3. Dictanta scopes the capture to the Webex process

Because audio capture is per-process, Dictanta lets you point it at the Webex Mac app (or the browser running the Webex web client) rather than at “anything making noise on this Mac.” If Webex is the only thing producing audio, that’s automatic. If you have Spotify playing or a YouTube video paused in another tab, none of that ends up in the transcript.

If you join Webex through the desktop app and a web client on the same Mac (some IT configurations push both), you can pick the source in Dictanta’s source picker before pressing record. Default is “the process with active audio.”

You also pick whether to include your microphone in the mix. For meetings the default is both: capture Webex’s audio (everyone else) plus your mic (you), and let SpeechAnalyzer transcribe the combined stream. For solo recordings — a voice memo, a dictated draft — flip to mic-only.

4. Live captions during the call

Two surfaces:

Menu-bar strip. Glanceable, about three lines, won’t appear in a screen share if you mute it.
Live transcript window. Open with ⌥⌘T. Larger, scrollable, useful when someone is reading out a contract clause or an account number and you want to verify you heard it.

Both update with about 300ms latency on an M-series Mac. The system-audio path itself is Mac-only — iOS and iPadOS don’t let third-party apps subscribe to other apps’ audio output, so the iPad and iPhone versions of Dictanta cover the mic-only and Voice Memo import paths, not the Webex system-audio capture.

5. Stop the recording, get a finished summary

When the Webex call ends — or when you press ⇧⌘R again, or click stop in the Dictanta window — three things happen locally:

Final transcript with per-segment timestamps. Each segment is tap-to-seek; clicking a line jumps the audio player to that moment.
Structured summary via Apple Foundation Models. TL;DR, decisions, action items (best-effort owner and due-date inference), open questions, follow-ups. The LLM runs on the Neural Engine — same chip path as the transcription — and never leaves the Mac.
Audio-anchored bullets. Every summary bullet links back to the audio span it came from. Click a bullet, the audio scrubs to that moment, the matching transcript line highlights. This is the part that separates verified summaries from plausible ones: when a bullet says “the customer agreed to a Q3 renewal,” you click it and hear whether anyone actually agreed.

6. Export, or just leave it

Dictanta v1.0 exports to Markdown, JSON, and plain text. Markdown drops cleanly into Notion, Obsidian, Bear, Apple Notes, Logseq, Craft. JSON is what you want for Shortcuts flows or your own automation. Plain text is for the simplest cases.

DOCX, PDF, and SRT export ship in v1.1.

Transcripts and summaries sync to your other Apple devices through CloudKit by default. The audio itself does not sync unless you turn on iCloud Drive backup explicitly — audio defaults to local-only with a seven-day auto-delete, both configurable in Settings → Privacy.

Webex-specific things worth flagging

A few quirks of Webex shape the flow slightly differently than Zoom or Meet.

Webex on Mac sometimes routes audio through its own virtual device. On older Webex desktop app versions and on some IT-managed configurations, the app installs a virtual audio driver. ScreenCaptureKit captures audio per running process regardless of which output device the process targets, so this doesn’t break capture — but if Dictanta’s source picker shows two processes (the Webex helper plus a virtual device endpoint), pick the helper. The helper is what’s actually generating speech.

Webex Personal and the 40-minute cap. Webex Personal limits meetings to 40 minutes when there are three or more participants. The cap is on the meeting itself, not on your ability to record it — Dictanta keeps recording continuously. When the meeting reconnects after the host restarts it, you can either keep the same recording running (one continuous file with a brief silent gap) or stop and start a new one. The former is usually what you want for action-item continuity.

Cisco AI Assistant transcripts vs. local transcripts. Webex AI Assistant produces transcripts in Cisco’s cloud, stored under the host’s Webex account, governed by Cisco’s data residency settings. Dictanta produces transcripts locally on your Mac, governed by nothing other than your own filesystem permissions. The two can coexist — nothing about local capture interferes with Webex’s own transcript if the host has it enabled.

Guest joins are the most common Webex case for this. Enterprise sales calls are disproportionately on Webex because the customer’s IT mandated it. As a vendor joining as a guest, you have no host permissions, no Webex Assistant access on your end, no path to the host’s transcript unless they email it to you. Local capture is the only route that lets a guest walk away with a transcript of the call they were just on.

End-to-end encrypted Webex meetings. Webex offers E2EE meeting modes for regulated verticals. The audio is still decrypted at your Mac in order to play through your speakers — that decrypted audio is what ScreenCaptureKit reads, and ScreenCaptureKit sees no plaintext anywhere other than where the OS already does. The capture path is the same as for a non-E2EE meeting; the E2EE only governs the wire, not what your Mac is playing.

Where the on-device path has limits

Honest list, same as in the Meet write-up:

No speaker tagging in v1.0. Webex Assistant labels speakers using the participant list — it knows which account is talking because Webex routes named audio streams. Dictanta on the audio-output path sees a single mixed stream. Diarization from a system-audio mix on short clips is on the v1.1 roadmap. If “who said what” is a must-have today, Webex Assistant (when you have it) or Otter’s calendar bot is the better tool.
You have to be at the Mac. Bots can sit in calls you’ve skipped. Local recording needs your Mac present, awake, and in the call.
No automatic Webex space push. Webex Assistant transcripts can post to a Webex space after the call. Dictanta exports to Markdown/JSON/plain text on your Mac. If you need the transcript automatically posted into a Webex team space, the native flow with Suite licensed is structurally easier.
AirPods bonded only to your iPhone. If you join Webex on the Mac but route audio to AirPods paired only to your iPhone, the audio bypasses the Mac’s audio output and there is nothing for ScreenCaptureKit to capture. Pair to the Mac (or use wired headphones, or the Mac’s built-in speakers) and the path works fine.

If those gaps don’t matter to your meetings, the on-device path is the cleaner default.

When Webex Assistant is the right call

Three cases where Cisco’s built-in transcription is genuinely the better answer:

Your organization standardizes on the Webex Suite for compliance reasons. If customer contracts require transcripts to live in Cisco’s certified data center under your enterprise’s Webex tenant, swimming against that current with a local recording costs more than you gain.
You need speaker labels today, on every call, with no setup. Webex’s transcription labels speakers automatically because of the account-level signal. Local recording can’t see that signal.
You’re recording every enterprise meeting at scale with retention policies. A company-wide policy of “all customer calls transcribed and retained for 7 years” is better served by the native flow with your enterprise’s retention controls applied. Local recording on each person’s Mac is a worse fit for org-wide compliance.

If you’re not in any of those buckets — and most contractors, vendors joining as guests, or individual contributors on Webex Meet without Suite aren’t — the local path is better.

What this unlocks that Webex’s native flow doesn’t

No Webex Suite upgrade required. Webex Suite is $10.50/user/mo on top of Meet. For a one-person shop on Webex Personal, or a small team on Meet without Suite, that adds up. The on-device path runs on the macOS 26 you already paid for.
Recording calls that aren’t yours. Joining a customer’s or vendor’s Webex as a guest gives you no transcription rights on their end. The local capture path is the only one that lets a guest walk away with a transcript at all.
No third-party cloud processor on the data path. Cisco’s own cloud is one thing. Adding Otter or Fireflies as a bot adds a second cloud processor on top. The local path adds zero new processors.
Sovereignty for federal and regulated work. A lot of Webex’s user base is in industries where adding any additional cloud vendor to the data path is a procurement hurdle. Local transcription removes the question.

Trying it on your next Webex call

Practical first-time flow:

Install Dictanta from the Mac App Store. Small download. The SpeechAnalyzer language model ships as part of macOS 26 and is already on your Mac.
Open Dictanta once. Grant microphone and screen recording permissions when prompted. Either accept the default hotkey ⇧⌘R or rebind it from Settings → Hotkeys.
On your next Webex call, press the hotkey when the call starts. Don’t think about it again during the meeting.
After the meeting, open Dictanta. The recording is in the meetings list labeled with the Webex source process, the timestamp, and the duration. Click into it for transcript and summary.
Export to Markdown if you want it in Notion or Obsidian, or leave it in Dictanta and rely on the on-device full-text search across all your recordings.

The free tier gives you three full recordings, no length cap, which is enough to compare the result against whatever Webex Assistant recap you would otherwise be reading (or not reading, if the host didn’t share it). Paid tiers are $9.99/mo, $79.99/yr, or $149.99 lifetime. No per-minute meter and no length cap at any tier — a three-hour quarterly business review transcribes the same way a ten-minute customer check-in does.

Bottom line

Webex’s native transcription is fine when you are the host, you have Webex Suite licensed, and your IT department lets you use it. For an enterprise user with the right plan, that path is the simpler answer.

For everyone else on Webex — guests on customer calls, Personal-plan users, Meet-plan users without the Suite SKU, anyone in a vertical where IT disabled Webex AI Assistant for compliance reasons — the local path is the missing option. Capture Webex’s system audio with ScreenCaptureKit, transcribe with SpeechAnalyzer, summarize with Apple Foundation Models, all on the Mac, no bot, no upload.

That is what Dictanta ships, the same way it ships the Zoom no-bot flow, the Teams no-bot flow, and the Meet on-device flow. If you are on Webex Suite and it’s working, keep it. If you’re not, the on-device path costs nothing to install and try on the next call.