Podcast Editor
Detects who’s speaking and cuts to their camera — including the wide shot when everyone talks at once.
What it does
Podcast Editor works out who is speaking when across your multi-camera recording and cuts the timeline so the active speaker's camera is always the one showing. If you recorded separate mic tracks, no AI is involved at all — it reads your mics directly, which is the fastest and most accurate path. Only have a mixed audio track? The built-in local speaker detection handles it free and offline, or plug in your own key for a cloud engine (Deepgram, AssemblyAI, Speechmatics or ElevenLabs).
The clever part: when two people talk over each other, it doesn't guess — it cuts to your wide shot, exactly like a human editor would. Pick a pacing style (Snappy, Balanced or Relaxed), preview every switch segment per speaker before applying, and optionally remove the silences in the same pass. Your original sequence is kept as a backup.
Key features
- No key needed — separate-mics mode uses zero AI, and mixed-audio mode ships with a free local engine. Cloud engines are optional, via your own keys.
- Wide-shot on crosstalk — simultaneous speech cuts to your group angle instead of ping-ponging between speakers.
- Three pacing styles — Snappy for shorts-style energy, Balanced for most shows, Relaxed for long-form conversation.
- Preview before apply — see the full switch timeline with per-speaker playback before your sequence changes.
- Silence removal in the same pass — tighten the episode while it switches, no second run needed.
- Original kept as backup — the untouched sequence stays in your project, one click away.
How to use it
Pick your audio source
Separate mic tracks (most accurate, no AI) or one mixed track (local engine, or a cloud engine with your key).
Map speakers to cameras
Tell PrimCut which speaker belongs to which camera angle — including your wide shot.
Choose the pacing
Snappy, Balanced or Relaxed — how eagerly the edit cuts between speakers.
Preview the switches
Review the segment timeline, play any speaker's moments, and adjust before committing.
Apply in one pass
Every camera switch (and optionally every silence cut) lands at once, with your original saved as a backup.
Watch the walkthrough
Frequently asked questions
Do I need an API key?
No. Separate-mics mode uses no AI at all, and mixed-audio mode includes a free local engine that runs offline. Cloud engines (Deepgram, AssemblyAI, Speechmatics, ElevenLabs) are optional upgrades via your own keys.
What happens when both people talk at once?
With separate mics or the local engine, overlapping speech is detected and the edit cuts to your wide shot — just like a human editor. That's a PrimCut speciality most cloud engines can't do.
Will it wreck my timeline?
No — you preview every switch before applying, and the original sequence is kept untouched as a backup in your project.
Can it remove silences too?
Yes. Tick “remove silences” and the dead air is cut in the same pass as the camera switching — one run, both jobs done.
Related tools
Get Podcast Editor in your panel.
Eleven tools, one Premiere Pro panel. Free tier, no credit card.