- Home
- AI Voice Isolators
- Gaudio DME Separation
Gaudio DME Separation
What it is
Gaudio DME Separation is an API service that isolates dialogue, background music, and sound effects from video. Extraction modes include dialogue, DME, music, M&E, and sound effects. Dialogue mode prioritizes speech clarity and is designed to separate spoken dialogue even when it overlaps with singing vocals, which Gaudio describes as a difficult audio-processing task. The service uses Gaudio Lab’s GSEP technology, and Gaudio Lab was founded in 2015. API access costs $0.24 per minute per track for all users. A higher-fidelity M&E model, gsep_dme_me2track_v2, is listed at $0.3600 per minute and described as cleaner and higher fidelity than me2. Gaudio also offers customized processing reviewed and optimized by audio post-production experts. Listed use cases include broadcasters, streaming and FAST platforms, independent creators, localization, dubbing, and copyright clearance. The product page labels its web service as coming soon for early 2026. Minor artifacts may occur in sound-effects extraction when speech or music resembles effects. Customer uploads and generated media are not used for model training without explicit consent.
For and against
For
- Multiple dialogue, music, and effects modes
- Dialogue extraction targets speech clarity
- API pricing is per track minute
- Optional expert-reviewed custom processing
- Uploads excluded from training without consent
Against
- Web service is still coming soon
- Sound-effects extraction may create artifacts
- Usage is billed per minute per track
The verdict
Gaudio DME Separation focuses on separating production audio through an API, with several modes and a higher-fidelity M&E option. Its per-minute pricing, upcoming web service, and possible effects artifacts are key considerations.
Who it suits
It suits media teams, creators, broadcasters, and localization or dubbing workflows needing separate dialogue, music, or effects tracks. API users can choose standard or higher-fidelity M&E processing.
Plans and prices
DME Separation API
API access; priced per track per minute
Read 7 Oct 2026 · gaudiolab.com
DME Separation API — higher-fidelity M&E v2
gsep_dme_me2track_v2; priced per minute
Read 7 Oct 2026 · gaudiolab.com
On the record
- API access
- The API service is available to all users at $0.24 per minute per track.gaudiolab.com · read 7 Oct 2026
- Company founding
- Gaudio Lab says it was founded in 2015.gaudiolab.com · read 7 Oct 2026
- Data and model training
- Gaudio’s privacy policy says customer uploads and generated media are not used for AI model pretraining or fine-tuning without explicit customer consent.gaudiolab.com · read 7 Oct 2026
- Dialogue and vocals
- Gaudio says its separation distinguishes spoken dialogue from singing vocals, which it describes as a challenging audio processing problem.gaudiolab.com · read 7 Oct 2026
- Dialogue extraction
- Dialogue extraction prioritizes speech clarity and is designed to separate dialogue even when it overlaps with vocals.gaudiolab.com · read 7 Oct 2026
- Expert processing
- Gaudio offers customized processing reviewed and optimized by audio post-production experts.gaudiolab.com · read 7 Oct 2026
- Higher-fidelity model
- The pricing page lists gsep_dme_me2track_v2 at $0.3600 per minute and describes it as cleaner and higher fidelity than me2.gaudiolab.com · read 7 Oct 2026
- Limitations
- The product page notes that minor artifacts may occur in sound effects extraction when speech or music resembles effects acoustically.gaudiolab.com · read 7 Oct 2026
- Modes
- Available extraction modes include dialogue, DME, music, M&E, and sound effects.gaudiolab.com · read 7 Oct 2026
- Purpose
- DME Separation automatically isolates dialogue, background music, and sound effects from video.gaudiolab.com · read 7 Oct 2026
- Technology
- The service is powered by Gaudio Lab’s GSEP audio separation technology.gaudiolab.com · read 7 Oct 2026
- Use cases
- The page identifies broadcasters, streaming and FAST platforms, independent creators, content localization, dubbing, and copyright clearance as use cases.gaudiolab.com · read 7 Oct 2026
- Web service
- The product page labels the web service as coming soon and says it will launch for all users in early 2026.gaudiolab.com · read 7 Oct 2026
