Seply
Splits mixed audio and video into separate time-aligned tracks per speaker, in 30+ languages — no manual cutting for podcasters
Last verified:
What is Seply?
Seply is an AI-powered speaker separation tool that automatically splits mixed audio and video recordings into individual, time-aligned tracks for each detected speaker. Designed for podcasters, interviewers, and content creators, it accepts files up to 1GB in 30+ languages and outputs editor-ready WAV files without manual cutting.
Seply pricing
Pricing model: Freemium
Free Trial (30 credits); Starter $12/mo (200 credits); Creator $29/mo (600 credits); Studio $65/mo (1500 credits)
Seply pros
- Time-aligned speaker tracks stay synchronized with the original recording
- Web-based with no installation; supports 30+ languages and multiple audio/video formats (MP3, WAV, MP4, MOV, etc.)
- Creates actual audio tracks per speaker (not just labels), enabling independent editing, muting, and enhancement
- Automatic speaker detection with optional manual override; 30-day private result access
Seply cons
- Free trial limited to 30 credits (≈3 minutes of separation); requires credit purchase for longer files
- File upload capped at 1GB; requires internet connectivity (no offline processing)
- Credit-based pricing system adds complexity to cost estimation across different usage types
Frequently asked questions about Seply
What audio and video formats does Seply support?
Audio: WAV, MP3, M4A, AAC, FLAC, OGG, OPUS, WMA, Speex. Video: MP4, AVI, MOV, MKV, WebM (processes original audio track without re-encoding).
How is speaker separation different from speaker diarization?
Diarization labels who spoke when; separation creates actual individual audio tracks for each speaker that you can edit, enhance, or reuse independently.
Is my audio data kept private?
Yes, Seply uses specialist processing infrastructure with 30-day private result access, and upload rights are required.
How many speakers can it detect?
Seply can detect speakers automatically or you can enter an expected number manually; no explicit limit is stated.