Rapid vs Local
Rapid Voice Cloning
Clones a voice quickly using just a few seconds of voice recordings. It is suitable for Rapid Prototyping and quick fixes of mistakes or mispronunciations in Voice Recordings.
Local Voice Cloning
Craft a voice clone using our StyleChanger model. Created from a few minutes to an hour of recordings, you can yield realistic, performative Hollywood-grade voice clones. The fine-tuning process runs locally on your computer, ensuring your privacy.


Compare Voice Cloning Options
Recording Requirements
Minimum
Minimum
5 seconds
2 minutes
2 minutes
Suggested
Suggested
20 seconds
30 minutes
30 minutes
Maximum (for best-in-class results)
Maximum (for best-in-class results)
40 seconds
1 hour
Infinite
Background Noise Handling
Background Noise Handling
Low
Medium
Excellent
Training Time
Average
Average
5 seconds
60 min
7 days
Quality
Intelligibility
Intelligibility
High
High
Excellent
Naturalness
Naturalness
Good
High
Excellent
Likeness
Likeness
Good
High
Excellent
Platform
Platform
Platform
Web
Desktop
Service
GPU Requirements
Minimum
Minimum
-
NVIDIA RTX 2080 / 3080 / 4080, 12 GB RAM
-
Suggested
Suggested
-
NVIDIA RTX 3090 / 4090, 24 GB RAM
-
Data Processing
Data Processing
Data Processing
Cloud
Locally (most secure)
Altered
⭐ Enterprise Custom Voices
Custom Voices, Tailored Services
For Hollywood-grade productions with demanding requirements, you can use Altered's Custom Voice services. This service captures the finest nuances of any voice, even from noisy recordings, allowing for the creation of content indistinguishable from the original performer.A dedicated account manager will guide you through the process, offering support for voice recording Quality Assurance, onboarding, troubleshooting, and any other needs you may have.

How it works
Frequently Asked Questions
Copyright © 2022-2023 Altered. All rights reserved.