Speaker Cut Wiki · complete
Every article, on one page.
The whole Speaker Cut knowledge base in one place: 31 articles, updated Nov 19, 2026.
Getting started
A Mac app that syncs your cameras and mics into a multicam, cuts it to whoever is talking, and sends a finished edit to Final Cut Pro.
Speaker Cut is a macOS app for people who edit talking-head shows in Final Cut Pro: podcasts, panels, interviews and live-switched shoots.
It does four jobs:
- Builds the multicam. Drop in camera files and audio-recorder files and Speaker Cut syncs them by sound. It can also import an ATEM ISO recording, or take a multicam you already built in Final Cut.
- Cuts it to whoever is talking. It listens to each person’s mic and switches the picture to them, with cuts that land just before the new speaker’s first word.
- Lets you watch and change the cut. The Cut tab plays the episode with every camera on screen. Press 1–8 while it plays to switch cameras, like a live switcher.
- Hands Final Cut a finished edit. One import brings the multicam, the cut, each mic on its own lane (with auto-mix keyframes), highlight reels and clips. Your original media stays linked.
With Speaker Cut One or Pro it also writes the transcript, AI show notes, highlights and cold-open hooks. Highlight Cut does all of that from a single recording with no separate mics.
Who it’s for
- Podcast and video-podcast editors working in Final Cut Pro.
- Teams that record panels or interviews with several cameras and wireless mics.
- Studios that record on an ATEM Mini Pro ISO or Extreme ISO.
- Anyone with one recording (Zoom, StreamYard, Riverside, a phone on a tripod) who wants clips and highlights.
Requirements
- A Mac with Apple silicon (M1 or later) running macOS 15.6 or later.
- Final Cut Pro 12 or later to finish and publish the edit.
- Full details: System requirements.
- No account is needed for the free app.
Your footage stays on your Mac
Syncing, the cut, transcription and auto-mix run on your Mac. Your video and audio are never uploaded. With Pro (or One with your own AI key) only the finished transcript text is sent to write notes and highlights.
Speaker Cut is made by Le Design & Media Co. The current version is 2.0.
Getting started
Apple silicon Mac (M1 or later), macOS 15.6 or later, Final Cut Pro 12 or later. Ideal – an M-series Pro or Max with 32 GB, footage on a fast SSD, and 100 GB free.
Minimum
|
Requirement |
| Mac |
Apple silicon – M1 or later. Speaker Cut does not run on Intel Macs. |
| macOS |
macOS 15.6 Sequoia or later. (The app itself needs macOS 15; Final Cut Pro 12 needs 15.6.) |
| Final Cut Pro |
Final Cut Pro 12.0 or later. Speaker Cut exports FCPXML 1.14, which Final Cut 11 and earlier can’t open. |
| Memory |
16 GB recommended. Speaker Cut holds each mic’s audio in memory while it works, so long episodes with many mics need more. |
| Disk space |
About 250 MB for the app, plus about 140 MB for the transcription model (downloaded the first time you transcribe), plus room for playback copies (see below). |
| Internet |
Only for downloading the transcription model the first time, activating a license, AI notes and highlights, and update checks. Syncing, the cut and the Final Cut export work offline. |
Ideal
- An M-series Pro or Max chip with 32 GB or more of memory for 6–8 cameras of 4K on long episodes.
- Footage on a fast SSD (Thunderbolt or USB-C NVMe). Reading every mic and making playback copies is much quicker than from a spinning drive or a NAS over Wi-Fi.
- 100 GB or more free on your internal drive for playback copies. Speaker Cut keeps up to about 80 GB of them and stops making new ones when less than 20 GB is free, so the Cut tab may stutter on a nearly full drive.
- The latest macOS and Final Cut Pro 12.x.
Playback copies
To play several 4K/HEVC/ProRes cameras smoothly, Speaker Cut makes light 960×540 copies in the background. Exports always use your original footage. See Smooth playback copies and storage.
Not supported
- Intel Macs – the app won’t open.
- macOS 14 Sonoma and earlier – the app won’t open.
- Final Cut Pro 11 and earlier – Final Cut can’t read the export. Update Final Cut from the Mac App Store.
Getting started
Download the free Mac app from getspeakercut.com, drag it to Applications, and Speaker Cut checks for updates each time it opens.
Download
- Go to getspeakercut.com and click Download free (or Download for Mac).
- Enter your name and email. The download (a
.dmg file) starts right away. You’ll also get the occasional tip, product update and subscriber-only discount; unsubscribe any time.
- Open the
.dmg and drag Speaker Cut into your Applications folder.
The app is signed and notarized by Apple.
First launch
The first time you open it, a short welcome walks you through three setup steps: footage access (letting Speaker Cut read footage on external drives and a NAS), choosing your plan, and creating your first show. See First launch and the tours.
Updates
Speaker Cut checks for updates when it opens and shows a banner when a new version is out. Download it from the site the same way and replace the app in Applications. Your projects and settings are kept. See what changed in each release on the changelog.
Requirements
An Apple silicon Mac on macOS 15.6 or later, and Final Cut Pro 12 or later. See System requirements.
Getting started
The welcome covers footage access, your plan and your first show. A five-step tour runs the first time you open a project; both can be replayed from Settings.
The welcome
On first launch Speaker Cut shows a welcome (“Your podcast, cut by who’s talking.”) with Get set up or I already have a license key, then three steps:
- Footage access – lets Speaker Cut read footage on external drives and the NAS. Allow it, or projects on those drives will show as offline.
- Plan – stay on Free, or activate One or Pro.
- Your first show – name the show and its hosts so voices and highlights start smarter. See Shows.
The project tour
The first time you open a project, a five-step walk-through points out the five tabs, playback, live switching, the cut settings and Export.
Replay either one
Both live in Settings: replay the welcome and its three setup steps, or replay the project tour (open any project and the tour starts there).
Getting started
Sync, the cut, transcription and auto-mix run on your Mac; video and audio are never uploaded. Only transcript text goes to the AI.
- Your footage is never uploaded. Syncing, speaker detection, the cut, transcription and auto-mix run on your Mac, and the cut works offline.
- AI features (notes, highlights, hooks, naming voices) send only the transcript text – to Speaker Cut’s hosted AI on Pro, or to your own Anthropic/OpenAI key on One. Never video or audio.
- API keys you paste on One are stored only on your Mac.
- Diagnostics are optional and anonymous: events and errors only, never your media, transcript, AI output or license.
- Speaker Cut never moves or changes your footage.
Getting started
Quick answers on Final Cut, multicams, live switching, one-recording workflows, internet and API keys, privacy, 4K playback and licensing.
Does it work with Final Cut Pro?
Yes. Speaker Cut hands Final Cut one import with the cut, your highlight reels and clips, and your original media stays linked.
Do I have to build the multicam in Final Cut first?
No. Drop camera files and recorders and Speaker Cut syncs them by sound, or pick an ATEM .drp. If you already synced one in Final Cut, export its XML and use I already have a multicam.
Do I need timecode?
No. Sync is by sound. Timecode is kept if your cameras have it.
Can I change the cut myself?
Yes. On the Cut tab, tap 1–8 while it plays to switch, or change pacing and auto-mix and the whole episode re-cuts. Your switches go to Final Cut.
What if I only have one recording with no separate mics?
Use Highlight Cut (One and Pro): it separates the voices, names who’s who, writes the transcript and finds highlights.
Do I need an internet connection or an API key?
Not for the free features: sync, the cut and the export are offline with no account. The transcript and AI are the paid layer – One uses your own Anthropic or OpenAI key; Pro is hosted with no key.
Is my footage uploaded?
No. Only the transcript text is sent for AI notes and highlights.
My cameras shoot 4K HEVC or ProRes. Will it keep up?
Yes – light playback copies are made in the background. Exports use your originals.
How many Macs can I use?
One active Mac at a time; activate a new Mac whenever you like and the license moves.
Which ATEM models?
ATEM Mini Pro ISO and ATEM Mini Extreme ISO recordings (the .drp project the ATEM writes).
How much does it cost?
Free forever for the editing engine; Speaker Cut One is $247 once; Speaker Cut Pro is $24/month.
What Mac and Final Cut version do I need?
An Apple silicon Mac (M1 or later) on macOS 15.6 or later, and Final Cut Pro 12 or later. Intel Macs and Final Cut 11 aren’t supported. See System requirements.
Starting an episode
Home lists your episodes by show and offers four starts – build from clips, import an ATEM recording, bring a multicam you have, or Highlight Cut from one recording.
Home shows your episodes grouped by show, with search, and four ways to start a new one. You can also drop footage anywhere on the Home window and Speaker Cut picks the right path.
| Start |
Use it when |
Guide |
| Build a multicam from clips |
You have camera files and/or separate audio recorders (lavs, wireless packs, a field recorder). |
Build a multicam from clips |
| Import an ATEM recording |
You recorded on an ATEM Mini Pro ISO or Extreme ISO and have its .drp project file. |
Import an ATEM recording |
| I already have a multicam |
You already synced a multicam clip in Final Cut Pro. |
Use a multicam you already have |
| Highlight Cut |
You have one recording with everyone in it and no separate mics. Needs One or Pro. |
Highlight Cut |
All four lead to the same episode with five tabs. See The five tabs.
Who’s who
The first three paths continue to Who’s who before the cut. Highlight Cut goes to its voices review instead.
Delivered episodes
When an episode ships, move it to the Delivered shelf. See The Delivered shelf.
Starting an episode
Drop camera and recorder files; Speaker Cut syncs every angle to a master by sound – no timecode needed – and joins takes a recorder split into pieces.
Start from Build a multicam from clips on Home (or drop the files on Home).
What to bring
- Every camera file and every external audio recorder file (lav packs, wireless mics that record onboard, a Zoom/field recorder, a StreamYard program track). Folders work too; lanes are named after their folders.
- Media stays where it is. Speaker Cut references it and never copies or uploads it.
Add the shoot
Drop the whole shoot folder, or use Choose a folder… / Choose files…. Each camera or recorder becomes one angle; files a recorder split into ~30-minute pieces are stacked into one angle. Rename an angle by clicking its name; click its file line to reorder, add or remove pieces.
Pick a master (◆)
Click the ◆ on the lane that should be the master. The master is the one feed that hears everyone – a program mix, a room mic, or a wide camera that hears the whole table. Every other angle is synced against it by sound.
- Remote recordings (StreamYard, Riverside): the program mix.
- In person with a recorder: the room mic or the recorder’s mix track.
- Cameras only: the wide.
Sync
Click Sync by sound. Speaker Cut lines everything up by audio, at many points across the session. Timecode isn’t required (if your cameras carry it, it’s kept). Each angle then shows Master, Locked · ‹offset›, Low confidence or No lock, and a banner says “Synced. All ‹n› angles line up with ‹master›.” Click Sync again after changes, or Next: who’s who → to continue (see Who’s who).
To change angles later, use Change angles on the episode’s Build tab, then Save to project.
It handles real-world messiness:
– Cameras and recorders that split a long take into pieces are joined end to end and play as one continuous feed.
– Files that started far apart still sync.
Think in people
Each person has a camera (the picture to cut to) and a mic (how Speaker Cut knows they’re talking). A webcam is both; an external recorder is audio-only; a wide or room camera is video-only and is used as the wide shot.
If an angle doesn’t sync
It didn’t share enough sound with the master. Usually the camera’s mic was off, or the master can’t hear that person. Give the angle scratch audio or choose a master that hears everyone, then build again. See Sync problems.
When it’s built, open the Cut tab – the episode is already cut. See The Cut tab.
Starting an episode
Pick the ATEM's .drp, choose a take, add your wireless-mic folders; cameras arrive in sync and each mic file is placed in the right take by its sound.
For shoots recorded on a Blackmagic ATEM Mini Pro ISO or Extreme ISO.
Steps
- On Home choose Import an ATEM recording and click Choose the .drp (it sits next to the Video ISO Files folder; you can also drop it on the window).
- Pick the take – each shows its length, start timecode and number of live switches; the longest take with cameras is picked for you. A take marked Program only has no camera files on disk (the ATEM only saved its program recording), so there’s nothing to cut between. Before 2.0.1, later takes in a multi-take session could wrongly show Program only even though every camera file was there – update Speaker Cut if you see that.
- Check the cameras – inputs that recorded black are left out; click a camera to leave it out yourself.
- Choose the master audio (see below).
- Click Add a folder (or Add files) for each mic recorder – wireless transmitters that record onboard (for example Hollyland or DJI packs) or a field recorder. Add the whole folder.
- Click Line everything up. The summary shows each recorder as Locked (or Partial) and any files set aside as Not in take. The aligned audio is saved in a Speaker Cut Audio folder next to the recording.
- Next: who’s who – see Who’s who.
What Speaker Cut does for you
- Every camera arrives in sync. Inputs that recorded only black are left out.
- The ATEM’s own live program cut is kept as a reference.
- Each mic file is found and placed by its sound: joined across the recorder’s ~30-minute chunks, lined up with the ATEM (typically within about a millisecond of the ATEM’s own mic audio), and set aside if it belongs to another take. The import summary tells you which files were set aside and why.
- It copes with a recorder that started before the ATEM did, and with a switcher mic input that had nothing plugged in (a silent input no longer throws off the matching – fixed in 2.0).
Master audio
If you ran a mixed feed into the ATEM – on a mic input like MIC 1, or into one of your cameras – pick it as the master audio when you import. It plays when auto-mix is off, comes into Final Cut as a muted backup lane, and Speaker Cut remembers your choice for the next recording. Every lane’s Audio menu also has a Master audio option.
Tips
- Keep each wireless pack’s files in its own folder named after the person (e.g. “H1 – Matthew”); lanes take the folder’s name.
- If a mic file isn’t placed in the take you expected, it may belong to a different take – check the import summary. See Sync problems.
Starting an episode
Export the multicam's XML from Final Cut Pro and drop the .fcpxml or .fcpxmld on Speaker Cut to go straight to the cut.
Already synced the multicam in Final Cut Pro? Use it.
- In Final Cut, build the multicam clip (select clips → New Multicam Clip…, sync by audio or timecode). Naming each angle by person and role (e.g. “April – Cam”, “April – Lav”, “Room – Wide”) makes setup quicker.
- Choose File › Export XML….
- In Speaker Cut choose I already have a multicam and drop the
.fcpxmld bundle, a zip of it, or the Info.fcpxml inside.
You land on the same cut as the other paths. The export package also includes your original multicam, untouched (Source.fcpxml).
Starting an episode
From a single .mp4, .mov or .wav, Speaker Cut separates the voices, names who's who, writes the transcript and finds highlights you can export to Final Cut. Needs One or Pro.
Highlight Cut is for one recording with everyone in it – a Zoom or StreamYard recording, a Riverside mixdown, a camera with a room mic, a phone on a tripod, or the edited episode. Video or audio-only (.mp4, .mov, .wav). It’s part of Speaker Cut One and Pro; on Free, Find who’s talking opens the upgrade sheet.
Have a separate mic for each person? Use Build a multicam from clips instead.
Step-by-step with screenshots: the Highlight Cut guide.
Steps
- On Home choose Highlight Cut and drop the recording (or click Drop one recording, or click to choose).
- Session name – the date and guest work best (e.g. “2026-09-14 · Sam Rivera”); the guest’s name is taken from it.
- Show – pick the show (or New show…). Its hosts are listened for and spelled as saved.
- People in it – Auto or 1–6. Choosing a show presets hosts + 1. If two people sound alike, pick the exact count.
- Click Find who’s talking. Speaker Cut separates the voices, transcribes and names who’s who on your Mac. A long episode takes several minutes; you can do something else meanwhile.
Check the voices
Each voice card shows the name, where it came from (Recognized voice, Host, Named from the talk, A guess. Check it), talk time and share, and up to three clips to play.
– Fix a name by typing over it – it saves automatically.
– Two people came out as one? Under Not quite right?, change the head-count and click Re-separate voices. It reuses the first listen, so it takes seconds; typed names are kept.
Then Continue to highlights. A Highlight Cut episode has Transcript, Highlights and Export tabs (no Build or Cut tab). See Highlights, notes and hooks and Export to Final Cut.
Privacy
Separating voices and transcribing happen on your Mac. Only the transcript text is sent to the AI – never your video or audio.
Starting an episode
Name each person, set whose mic each audio lane is and what each camera shows, and pick one wide. Continue to cut unlocks when everyone has a mic and there's a wide.
After building a multicam, importing an ATEM recording or bringing a Final Cut multicam, Speaker Cut asks who’s who. It’s also on the episode’s Build tab (button Save changes).
Show
Pick the episode’s Show. The cut feel and auto-mix start from that show’s settings. If the show knows its hosts’ voices, Speaker Cut listens for them and offers “Recognized by voice from ‹show›…” with Use these names.
People
Name everyone who talks (Add a person for more). Each person needs a mic; a camera is optional.
Mic column – whose audio is each lane?
Set each recorder to its person (‹Name›’s mic), the feed that hears everyone to Master audio, anything else to No mic. Each person has one mic. Listen plays a few seconds where that person talks (Another spot for more).
Shows column – what does each camera show?
A person, a pair (A + B, a two-shot), All speakers (wide), or Don’t show. Exactly one camera must be the wide. Audio-only lanes say Audio only.
Two or more cameras on one person: one is their A-cam, the others are B-cams that punch in now and then for variety (click B-cam · make A to swap).
Continue to cut
Continue to cut stays inactive until every person has a mic and one camera is the wide; the status line says what’s missing (“Pick the wide shot.”, “‹Name› needs a mic.”). Until who’s who is done, the Cut, Transcript, Highlights and Export tabs send you back to Build.
Inside an episode
Every episode has Build, Cut, Transcript, Highlights and Export, in one window.
Open any episode and you get five tabs, left to right:
| Tab |
What it’s for |
| Build |
Makes the multicam: your cameras, mics, the master and who’s who. Relink moved footage here. |
| Cut |
Watch the cut with every camera on screen, switch it live with 1–8, and tune pacing, the wide and auto-mix. See The Cut tab. |
| Transcript |
Read and fix the transcript (One and Pro). See Transcript. |
| Highlights |
Notes, highlights and cold-open hooks (One and Pro). See Highlights and hooks. |
| Export |
Send everything to Final Cut in one import, or build a package with PDFs, captions and chapters. See Export to Final Cut. |
The top bar also shows the episode name, its show, and its status (for example “Cut ready · transcribed · auto-mixed”), with Export to Final Cut always one click away.
Inside an episode
The episode is already cut when you open it. Press Space to play, 1–8 to switch cameras like a switcher, ⌘Z to undo; every setting change re-cuts the whole episode.
When you open the Cut tab, the episode is already cut: Speaker Cut followed whoever was talking. Press play and watch – the switching you see is exactly what lands in Final Cut.
What’s on screen
- Program view – the camera that’s on air, with the timecode.
- Camera tiles – every camera (up to 8), numbered, with the one on air marked.
- Transcript (left) – follows the playhead.
- Inspector (right) – Overview (each person’s share of the screen vs. their share of the talking), Cut feel and Auto-mix.
- Timeline – the cut laid over each person’s mic lane.
Keyboard
| Key |
Does |
| Space |
Play / pause |
| ← → |
Back / forward a second (hold ⇧ for 5 seconds) |
| 1–8 |
Cut to that camera |
| ⌥ 1–8 |
Switch the whole shot |
| ⌘Z |
Undo the last switch |
Switch it yourself
Disagree with a shot? Tap the camera’s number while it plays, like a live switcher. Your switch replaces the cut from the playhead to the next change. Your switches are kept when the episode re-cuts and go to Final Cut with everything else.
Change how it cuts
The inspector has three tabs:
– Overview – On screen vs. talking for each person, shot count, your switches, average shot, Undo and Reset cut.
– Cut feel – Pacing (Calm, Balanced, Lively), Shortest shot, Ignore a quick “yeah”, Quick interjections (Stay on speaker / Cut to them), Go wide on cross-talk (Off, 3s+, 2s+, Any – One and Pro, needs a transcript), Cut back to the wide (Never … Lots), plus More options (B-cam variety, wide length, longest on one person, speech sensitivity).
– Auto-mix (One and Pro) – on/off, Turn quiet mics down (Gentle, Balanced, Aggressive, Gate), fades; drag a ◆ on a mic lane to reshape it.
Every change re-cuts the whole episode in a moment, keeping your switches. Save for ‹show name› makes your settings the default for every new episode of that show (the episode needs to be in a show).
See also How cuts are timed and Auto-mix.
Inside an episode
Cuts land just before the new speaker's first word and between words; quick interjections don't flip the picture; real cross-talk can hold the wide.
Speaker Cut reads the words, not just the audio levels, so the picture changes where an editor would change it.
Just before the first word
A cut lands a beat before the new speaker starts, and between words – not half a second into their sentence (new in 2.0).
Stays on whoever is talking
When someone gets a few words in mid-answer (“Wow.”, “Right.”), the picture stays with the person telling the story instead of cutting over and straight back. Turn it off in Settings › Cut (“Stay on the speaker through quick interjections”).
Go wide when people talk over each other
When two people genuinely talk at once, the cut can hold the wide (or their two-shot) for at least 3 seconds instead of bouncing between faces. Speaker Cut tells real cross-talk from mic bleed using the transcript (the same words on two mics is bleed; different words is cross-talk). Choose Off, 3s+, 2s+ or any overlap. Needs a transcript, so it’s part of One and Pro.
The wide as a breather
You can also let the wide come in now and then as a breather shot of the whole room. Off lets the talking decide.
Pacing
Pacing (how long shots last and how eager switches are) is set per episode on the Cut tab, or as a default in Settings. Save it for a show to reuse it on every episode.
Inside an episode
Each person's mic opens when they talk and ducks when they don't, as ordinary Final Cut volume keyframes on its own lane, plus a muted backup of the full mix. One and Pro.
With Auto-mix on (One and Pro), each person’s mic comes into Final Cut on its own lane with volume keyframes that open it when they talk and turn it down when they don’t – mic bleed and room tone gone, reactions kept.
- Nothing is baked in: they’re ordinary Final Cut keyframes you can drag.
- How far to turn quiet mics down: more keeps the room cleaner; less keeps more natural reactions.
- You can reshape each mic’s keyframes on the Cut tab.
- Auto-mix off: every mic stays open and you mix it in Final Cut.
Other audio options (Settings › Audio)
- Level matching – brings every voice to the same loudness on export, kept below clipping. Leave it off if you level in Final Cut.
- Backup mix lane – adds the room or ATEM mix (your master audio) underneath the mics, muted, just in case.
- Audio roles – one per person (a Speakers role with each person’s name, so you can solo or mute them in the timeline index) or one Dialogue role for all mics. The backup carries a Master Audio role.
Inside an episode
A word-by-word transcript credited to whoever said each word. Reassign speakers, search, make highlights from a selection, and download .srt, .txt or Word. One and Pro.
With One or Pro, the Transcript tab writes a word-by-word transcript on your Mac. On Free the tab is locked.
Transcribe
Click Transcribe on the Transcript tab. On in-person shoots every mic hears everyone, so Speaker Cut works out who said each word – each line appears once, credited to the right person.
Fix who said a line
Click the speaker’s name ▾ and choose from Who said this? The paragraph is marked reassigned, and captions, the export and the AI all use the change.
Find, listen and select
- Find in transcript (⌘F), with ↑/↓ between matches. Follow playback keeps the text in step with the video.
- Click any word or timestamp to jump the player there; Space plays and pauses.
- Click a speaker in the Speakers panel to show only them; Show everyone undoes it.
- Show in the text: Talking over each other, Pauses longer than 1 second, and Overheard copies (lines removed because another mic picked them up).
- Select text to Make highlight or Copy with timecodes.
Download
Captions .srt, Transcript .txt or a Word doc – saved to the episode’s export folder and shown in Finder. The export package can also include Transcript.pdf and Transcript.srt.
Inside an episode
Find highlights that tell a whole story and cold-open hooks, get YouTube titles and thumbnail text, steer them with a direction, re-roll any clip with a note, ask for more by type, and build reels. One and Pro.
The Highlights tab (One and Pro) finds the episode’s best moments from the transcript: highlights, cold-open hooks, and chapters from the show notes. On Free the tab is locked.
Find highlights
Optionally type what to look for in the box (e.g. “stories about the early days”), then click Find highlights (or Transcribe and find highlights if there’s no transcript yet).
Every highlight is a whole story
A setup, a turn and a payoff, stitched from anywhere in the episode, starting and ending on whole sentences, ending on its payoff. Each card shows its length, how many pieces it’s stitched from, its story type (3-act story, Thought process, Core principle, Callback) and a quote. Why these picks? explains what the AI looked for.
Refine one
Open a highlight to see each piece. ↻ Re-roll opens Tell the AI what to change: write a note (“tighter”, “add the part about pricing”) and Re-roll this clip, or leave it empty and Find a different one. Very short pieces are flagged short & unclear · re-roll?.
Ask for more
Find another highlight / Find another hook lets you choose what to look for – Surprise me, A story, A callback, A core principle, How they think, Advice, A hot take, Something funny – or type your own, ▶ Preview each, and add the ones you want. From a chapter: + Highlight from this chapter or + Hook.
Reels
+ Add to reel and + Add to cold open build the Highlights reel and the Cold open; reorder them on the Reels tab. The show learns your taste from the picks you keep.
Titles and thumbnail text
Since 2.0.1, the episode’s notes come with 5 YouTube titles and 5 thumbnail headlines (one or two words, in capitals), written from what was actually said – names, numbers and moments from this episode, not generic filler. Open Highlights › Titles:
– Click a title or headline to copy it. Each title shows its length; anything over 70 characters (where YouTube starts cutting titles off) is marked.
– New ideas writes a fresh set. Type a steer first if you like (“lead with the $2M mistake”, “more curiosity, less hype”).
– Earlier ideas keeps every set you’ve had, so nothing is lost when you ask for more.
– Episodes made before 2.0.1 show Write titles – one click adds them without redoing the notes.
The titles and headlines also open the Notes & Highlights PDF.
Getting them out
The reels and a clip per highlight and hook go to Final Cut in the same import. See Export to Final Cut.
Privacy
Only the transcript text is sent to the AI (Pro’s hosted AI, or your own key on One) – never video or audio.
Inside an episode
One import brings the multicam, the cut, mic lanes, highlight reels and clips into Final Cut from your original media; the package adds PDFs, captions, chapters and stills.
Send to Final Cut
Export to Final Cut (top right of every tab) opens the Export tab. Send to Final Cut builds the timeline from your original footage (never the playback copies), with each mic on its own lane. One import brings:
- the multicam and the cut,
- each person’s mic lane (with auto-mix keyframes on One and Pro) and the muted backup mix,
- the Highlights reel, the Cold open reel and a clip per highlight and hook, each its own project.
Final Cut opens and asks which library to add the event to. If it doesn’t open, Speaker Cut shows the saved file – import it with File › Import › XML…. Needs Final Cut Pro 12 or later.
On the Export tab you can also choose Auto-mix keyframes, Level matching and a Backup mix lane, and audio roles One per person or One Dialogue role.
The package
Build package (.zip) builds a package instead. The Final Cut file and a READ ME always go in; tick what else you want:
| File |
What it is |
Final Cut.fcpxml |
The import for Final Cut |
| READ ME |
How to import, for whoever opens it |
Notes & Highlights.pdf |
Summary, show notes, highlights, hooks |
Transcript.pdf |
Every word, by speaker |
Transcript.srt |
Captions for YouTube or Final Cut |
YouTube chapters.txt |
Chapters, ready to paste into the description |
| Camera stills |
A still from each camera |
Source.fcpxml |
Your original multicam, untouched (when you started from one) |
Transcript, notes and chapters need One or Pro.
Where it’s saved
By default, next to the episode’s footage when that drive can be written to. Change… on the Export tab picks another folder for that episode; Settings can send packages to Downloads instead.
Inside an episode
Put each episode in a show; it remembers hosts' voices, spelling, cut feel, auto-mix and the highlights you liked.
A show groups episodes on Home and remembers how you like them:
- Hosts’ voices – who’s who fills itself in on the next episode (One and Pro).
- Names and spelling of your hosts in the transcript.
- Cut feel and auto-mix – click Save for your show on the Cut tab to make your settings the default.
- Highlights you liked – ratings steer future picks (One and Pro).
- Tone, audience and topics, which guide the AI.
Every show keeps its cut style on Free too. Create one during first launch, from Home, or when starting an episode.
Inside an episode
Move a shipped episode off your projects with one click; search still finds it and one click brings it back.
When an episode ships, click Move to Delivered on its card (episodes that have been cut). It leaves Projects so Home stays tidy; the toast offers Undo. Search still finds it, and Back to Projects brings it back. Nothing is deleted.
- Open the shelf from the ‹n› delivered link at the bottom of Home.
- If several cut episodes haven’t been touched for two weeks, Home offers “‹n› older episodes look finished” with Move ‹n› to Delivered.
- A Delivered ‹date› badge means the episode was exported.
- Separately, the trash icon moves a project to Recently deleted for 30 days (Restore or Delete forever).
Settings & storage
Speaker Cut makes light playback copies of heavy 4K/HEVC/ProRes files in the background; exports always use your originals. See or turn them off in Settings › Storage.
Heavy camera files (4K, HEVC, ProRes, log) can stutter when several play at once. Speaker Cut makes light playback copies in the background so six cameras play smoothly on the Cut tab.
- Exports and Final Cut always use your original footage.
- Your footage is never moved or changed.
- Settings › Storage shows the space the copies use and what Speaker Cut keeps to open projects fast. Turn off Make smooth-playback copies to play your original files instead (which can stutter with several 4K cameras).
- Each recording’s audio is read once and remembered, so re-opening a project is instant.
Settings & storage
What lives in each Settings section – plan & AI, cut defaults, audio, Final Cut export, voices/shows, storage and about.
| Section |
What’s there |
| Plan & AI |
Your edition and license key; what powers transcripts, notes, highlights and hooks (Pro’s cloud AI, or your own Claude/GPT key on One – keys stay on your Mac). |
| Cut |
Defaults for new projects: pacing, the wide, “Go wide when people talk over each other”, “Stay on the speaker through quick interjections”. Each episode can change them on its Cut tab. |
| Audio |
How each mic lands in Final Cut: auto-mix, how far to turn quiet mics down, level matching, backup mix lane, audio roles. |
| Final Cut |
What you get when you export, where it’s saved, and “Open Final Cut after sending”. |
| Voices / Shows |
Each show’s cut style; recognizing voices and learning from the highlights you like (One and Pro). |
| Storage |
Smooth-playback copies, the space they use, drafts that were never cut. |
| About |
Updates (checked when the app opens), footage access for external drives and the NAS, replay the welcome or the project tour, and anonymous diagnostics (events and errors only – never your media, transcript, AI output or license). |
Plans & licensing
Free is the full editing engine forever. One ($247 once) and Pro ($24/month) add the transcript, AI, Highlight Cut and auto-mix.
Speaker Cut comes in three editions. Every edition includes the multicam builder, the speaker cut and the Final Cut export.
|
Free |
One |
Pro |
| Price |
$0, forever |
$247 once |
$24 per month |
| Multicam from clips or an ATEM recording |
✓ |
✓ |
✓ |
| Auto speaker cut you can switch live |
✓ |
✓ |
✓ |
| Export to Final Cut |
✓ |
✓ |
✓ |
| Full transcript + SRT captions |
– |
✓ |
✓ |
| AI show notes, highlights and hooks |
– |
✓ |
✓ |
| Highlight Cut (from one video) |
– |
✓ |
✓ |
| Auto-mix and deep clean (no mic bleed) |
– |
✓ |
✓ |
| How the AI runs |
– |
Your own Anthropic (Claude) or OpenAI key |
Hosted by Speaker Cut, no key needed |
| Licensed for |
– |
One Mac at a time |
One Mac at a time |
Free
The full editing engine with no account and no key. Download it from getspeakercut.com.
Speaker Cut One ($247, one-time)
Unlock the transcript and AI for good. One uses your own Anthropic or OpenAI API key, which is stored only on your Mac. The app also mentions Apple’s on-device AI on macOS Tahoe as an option for One. Buy it at getspeakercut.com/product/speaker-cut-one/.
Speaker Cut Pro ($24/month)
Zero setup: Speaker Cut hosts premium cloud AI for you, so there’s no key to paste and it works on any supported macOS. Pro includes a generous monthly generation quota. Subscribe at getspeakercut.com/product/speaker-cut-pro/.
Which one should I pick?
- Only need the multicam and the cut? Free.
- Want transcripts and AI, and already have (or are happy to get) an Anthropic or OpenAI key? One – pay once.
- Want it to just work with no keys? Pro.
Both One and Pro unlock the same features: the difference is how you pay and where the AI runs.
Plans & licensing
One and Pro licenses are active on one Mac at a time. Activate in Settings; release a Mac to move the license.
Speaker Cut One and Pro are licensed for one active Mac at a time, and moving is self-service.
Activate
- Buy One or start Pro on getspeakercut.com. Your license key is shown on the order confirmation page, in your order email, and on the order in My account on getspeakercut.com.
- Open Speaker Cut, go to Settings, and paste the full key into the plan section.
If the app says “That doesn’t look like a full key”, copy the whole key from the email again (it’s easy to miss a character).
Move to another Mac
- In Settings on the old Mac, click to release it (click again to confirm). The app then says “This Mac is released. Activate the key on your other Mac.”
- Activate the same key on the new Mac.
- You can also simply activate on the new Mac; the license moves over.
The free app
The free edition needs no account and no key.
Pro subscription status
- If you cancel Pro, access continues until the end of the period you’ve paid for.
- If a renewal payment fails, your license stays active while the payment is retried.
- When a subscription ends or expires, the AI features turn off. The free features (multicam, cut, Final Cut export) keep working.
Coupons
Discount codes are entered at checkout.
Troubleshooting
Report a bug, suggest something or ask a question from inside the app – Help menu, Home, Settings, or the Report this button on any error screen. It goes straight to the team.
Since 2.0.1 you can reach the Speaker Cut team from inside the app. Your message goes straight to the team, and we reply by email.
Ways in
- Help menu › Report a Bug… or Send Feedback…
- The Feedback link on Home.
- Settings › Help and feedback.
- Report this on any error screen – it attaches the error message for us, so you don’t have to copy it.
Sending it
Pick what it is – Something’s wrong, Idea, Question or Love it – write a few words, and click Send feedback (or press ⌘ Return). For a bug, say what happened and what you were doing right before; steps help a lot. If you have a license, we reply to your license email; on Free, add the email you’d like us to reply to.
What’s sent
Your note and its category, the app and macOS version, which screen you were on, the last few in-app steps (so we can see what led up to it), and the error text if you used Report this. Never your footage, audio or transcripts.
Troubleshooting
What to do when an angle won't sync, a wireless-mic file isn't placed, or files are set aside on an ATEM import.
An angle won’t sync (“No lock”)
Speaker Cut syncs by sound against the master. If an angle didn’t share enough audio:
– Check that angle actually recorded sound (camera mic on).
– Make sure the master hears everyone – a program mix, room mic or wide camera. Switch the master and build again; it only takes seconds.
A wireless-mic file wasn’t placed (ATEM import)
- The file may belong to another take – Speaker Cut sets those aside and the import summary says which and why.
- Add the whole folder from each recorder, not single files; files split into ~30-minute chunks are joined automatically.
- A recorder that started before the ATEM is fine.
- Pick the right master audio (e.g. the ATEM input carrying your mix, like MIC 1).
- Fixed in 2.0: a switcher mic input with nothing plugged in no longer stops wireless mics from syncing.
Files set aside on import
Files from another take, or the ATEM’s own audio files, are set aside automatically instead of stopping the import. The import summary lists them.
Still stuck?
Use the chat on getspeakercut.com and tell us the shoot setup (cameras, recorders, ATEM model) and what the import summary says.
Troubleshooting
Let the smooth-playback copies finish, keep them turned on, and make sure the footage drive is fast and connected.
- Let the copies finish. Right after import, Speaker Cut is still making light playback copies in the background. Playback gets smooth as they finish.
- Keep copies on. Settings › Storage › Make smooth-playback copies. Off means the Cut tab plays the original 4K/HEVC/ProRes files, which can stutter with several cameras.
- Slow or busy drive. Footage on a slow USB drive or a busy NAS can hold up the first play while files are read.
- Exports are unaffected – Final Cut always gets your original footage.
Troubleshooting
Final Cut opens and asks for a library; otherwise use File › Import › XML on Final Cut.fcpxml. Media links to your originals.
- You need Final Cut Pro 12 or later. Final Cut 11 and earlier say the file’s version isn’t supported – update Final Cut from the Mac App Store.
- With Open Final Cut after sending on, Final Cut opens and asks which library to import into. Pick one and the event appears with the multicam, the cut, reels and clips.
- Otherwise Speaker Cut reveals Final Cut.fcpxml in Finder. In Final Cut choose File › Import › XML… and pick it.
- Media links to your original files, so keep the footage drive connected when importing.
- Each mic is on its own lane with roles (one per person, or one Dialogue role) and the muted backup mix underneath.
- To re-import after changes, export again; you’ll get a fresh event.
Release notes
Speaker Cut 2.0.1 (Nov 19, 2026) – titles and thumbnail text, Report a Bug from anywhere, ATEM \"program only\" fix. 2.0 (Nov 12) – one window, live switching, five tabs, Shows, a Delivered shelf, editor-timed cuts, smooth 4K playback.
Latest: Speaker Cut 2.0.1 – released November 19, 2026. Full history: changelog.
2.0.1 (Nov 19, 2026)
- New – Titles and thumbnail text. Five YouTube titles and five one- or two-word thumbnail headlines for every episode, on Highlights › Titles and in the Notes & Highlights PDF. See Highlights, notes and hooks.
- New – Report a bug from anywhere. Help › Report a Bug… / Send Feedback…, a Feedback link on Home, and Report this on any error screen. See Report a bug or send feedback.
- Fixed – ATEM takes showing “Program only.” Later takes in a multi-take ATEM session now show all their cameras, with the right frame rate.
- Fixed – the Help menu’s guide and wiki links open properly.
2.0.0 (Nov 12, 2026)
New
- A whole new Speaker Cut. Everything happens in one window. Home shows episodes by show, with four ways to start. Drop footage anywhere on Home.
- Watch the cut, and switch it live. Every camera on screen, mics mixed the way Final Cut will hear them; tap 1–8 while it plays. Changing pacing or auto-mix re-cuts the episode in a moment.
- Transcript, Highlights and Export, side by side in five tabs: Build, Cut, Transcript, Highlights, Export.
- Shows remember hosts’ voices, cut feel, auto-mix and highlights you liked.
- A Delivered shelf for shipped episodes.
Improved
- Cuts that land like an editor’s – just before the first word, between words; quick interjections don’t flip the picture (toggle in Settings).
- Smooth playback with 4K, HEVC and ProRes via background playback copies; exports use originals.
- Faster import – mics read about twice as fast, each recording read once, re-opening is instant.
- Building a multicam – split takes work end to end, far-apart files still sync, lanes named after folders, and wireless mics sync reliably on ATEM shoots with an empty switcher mic input.
- Highlights – every highlight is a complete moment ending on its payoff.
Earlier highlights
- 1.8.3 (Nov 1, 2026): camera stills on the Cut tab, choose your master audio, muted backup mix lane in Final Cut, ATEM import fixes.
- 1.8.0 (Oct 20, 2026): ATEM recording import, go wide on cross-talk, cleaner in-person transcripts, sentence-accurate hooks.