Speaker Splitter separates multi‑speaker recordings into individual speaker tracks, directly inside your DAW. Everything runs locally — no cloud uploads, subscriptions, or external processing.
From $29 · Perpetual license — buy once, use forever, with free minor updates.
Works in any VST3 host — fully tested in
On‑device AI listens to the whole file, works out who is speaking when, and writes one clean audio file per person — next to your original recording.
Load Speaker Splitter in any VST3 host and browse to your recording. Pro accepts WAV, MP3, AIFF, FLAC and OGG.
Tell it how many people are in the recording — or let Pro's Auto mode detect the speaker count for you.
The AI analyses the file and writes one stem per speaker next to your original — plus SRT subtitles and a CSV timeline in Pro.
Isolating each voice by hand is slow. Speaker Splitter does the tedious part with on‑device AI, so you keep full control of the mix.
Speaker segmentation plus neural speaker‑identity embeddings decide who speaks when — no cloud API behind it.
No upload, no account, no internet needed to process audio. Client interviews and unreleased material stay on your machine.
Export per‑speaker WAV stems, ready‑to‑use SRT subtitles and a CSV diarization timeline in a single run.
Auto hardware mode speeds up analysis on GPU — CoreML on macOS, DirectML on Windows — with automatic CPU fallback.
Basic handles up to 10 speakers; Pro goes to 20 with Auto detection plus sensitivity, min‑segment and merge‑gap controls.
Windows 10+, macOS 13.3+ (Intel & Apple Silicon) and Linux. Tested in Audacity, REAPER, Live, Logic, Cubase and more.
Two unedited runs — a two‑person interview split in Audacity with Basic, and a full stems‑plus‑subtitles export with Pro.
Pick the WAV, set the speaker count, run — one clean track per voice appears next to the original recording.
Auto speaker detection, GPU‑accelerated analysis and a multi‑export workflow: audio stems, subtitles and a timeline in one pass.
It helped rescue a project we thought would have to be discarded.
Local processing was important because the recordings contained sensitive client information.
Both editions do the same core job — splitting speakers, fully offline. Pro adds formats, speed and export power for serious workloads.
The essentials — perfect for straightforward podcast and interview splits.
The full toolkit for post‑production, transcription and research workflows.
Every purchase is a perpetual license — pay once, use forever, with free minor updates.
| Capability | Basic | Pro |
|---|---|---|
| Price | $29 one‑time | $119 one‑time |
| Input formats | WAV | WAV, MP3, AIFF/AIF, FLAC, OGG |
| Max speakers | Up to 10 | Up to 20 + Auto detection |
| Processing hardware | CPU | GPU Auto mode + CPU (Win/macOS) |
| Exports | WAV stems | WAV stems + SRT + CSV |
| Fine‑tuning controls | Speaker count | Sensitivity · min segment · merge gap |
| Fully offline & private | Yes | Yes |
| Platforms | Windows 10+ · macOS 13.3+ (Intel & Apple Silicon) · Linux — VST3 | |
Every second of analysis runs on your own computer. There is no cloud upload, no account, and no internet connection required to process audio — which makes Speaker Splitter safe for client work, unreleased material, confidential interviews and research data.
If Speaker Splitter Basic or Pro cannot be installed, activated, or used in a supported operating system and DAW, contact us within 14 days of purchase. We will first help troubleshoot the issue; if the problem cannot be resolved, we will issue a refund.
Refunds may also be issued when a listed feature materially fails to work as described.
Because speaker‑separation results depend on recording quality, noise, overlapping speech, microphone setup, and other source‑audio conditions, refunds are not normally provided solely because the output differs from a buyer's expectations.
Refunds are not provided for change of mind after successful installation and use, except where required by applicable law.
Purchases made through MuseHub are processed according to MuseHub's refund procedure.
Let on‑device AI put every speaker on a separate track — and get back to the creative work.
Windows · macOS · Linux · VST3 · From $29, perpetual license
Speaker Splitter