Altered builds speech technology that separates the identity of a voice from the performance behind it. The platform is split into two clearly different products. Altered Studio is aimed at media production: it lets editors morph a recorded performance into any voice in a curated portfolio or a custom voice, clone a speaker from a short sample, generate natural text-to-speech narration and clean up noisy dialogue. Because it works after recording, it can use high-end GPUs and prioritise fidelity over speed. Altered Real-Time Pro is the live counterpart, using only a few CPU cores to transform a voice within milliseconds of it being spoken, which makes it usable in voice calls, streaming and games.
The real-time product is organised around job-specific feature sets rather than a single generic slider. Voice Skins let gamers and streamers take on a completely different vocal identity, including accent changes. Accent Translation is aimed at contact centres that want an American or British accent on a call without discarding the agent's own voice. Isolate suppresses background noise and competing speakers so only the intended speaker is captured. Euphonia targets voice restoration, easing dysphonia and disfluencies for people whose voice no longer matches how they want to sound.
Beyond the self-serve tools, Altered offers premium custom voice cloning work, where the company creates a specific voice for a project. Altered publishes an ethics page and states that consent is required for cloning real people, which matters for anyone using the tools commercially. Pricing is freemium: a free tier and trial are available, paid plans are sold for the software, and custom voice work and enterprise licensing are quoted after contact.
