Speech synthesis with regional accents, emotion and custom voice cloning
CereProc is an Edinburgh speech technology company that has been building text-to-speech voices since the mid 2000s, known for character and regional accents rather than flat narration. The CereVoice engine runs as an SDK on desktop and mobile, as a server for concurrent streams, and as a cloud API billed per character, and its voices are used for game characters, announcements, education and assistive communication. Custom voice work covers a service for building a copy of your own voice and studio projects such as reconstructing a voice for someone losing theirs. Much of the business is licensing, so talks start with a quote.
Last updated: 2026-09-20. This site only provides an index; for exact features, pricing, and licensing, see the official website.
game characters, accessibility tools, announcements and regional accent narration
If you're comparing similar products, check the alternatives below, or browse all tools in the AI Avatars & Digital Humans category.
Accent and character. The voice library covers regional variants and emotional delivery that large generic providers often skip, which is why games and accessibility projects use it for characters rather than neutral narration.
Yes, the CereVoiceMe service builds a personal voice from recordings made at home, and the company also works on voice reconstruction for people losing their speech to illness, using recorded material and adaptive techniques.
There is a cloud signup with a limited monthly allowance for developers, and academic use of the SDK is free. Commercial deployments are licensed, usually after a discussion about volume and target platform.
Animate a still photo into a talking presenter or a live agent
Turns one image plus audio or text into an expressive talking character
Video translation and AI avatars for speaking content
Speech synthesis for devices, accessibility and brand voices