CereProc Unveils Innovative AI Speech Software Update, Boosting Speed on all Devices

Edinburgh, Scotland – CereProc Ltd., a leading innovator in synthetic speech, is thrilled to announce a significant update to the CereWave AI deep neural network: Version 7. With CereWave AI V7, users will gain remarkable speed improvements, particularly on low spec devices such as mobile phones and tablets, improved prosody, and superior pronunciation in many languages, including multi-lingual word embedding support.

In response to the growing demand for even higher quality and faster synthetic speech across a wide range of devices, CereProc’s team of engineers has been diligently working, optimising the performance without compromising the quality of our DNN CereVoices. CereWave AI V7 represents a major milestone in CereProc’s ongoing commitment to delivering high quality synthetic speech to all and includes a large number of software improvements.

Key highlights of the software update include:

• Speed: All our CereWave AI CereVoices have had their generation speed increased, allowing for audio to be created faster especially on older, lower technical specification devices.
• Cloud access to Voice Genres: With V7, CereProc has now provided access to our Voice Genre system via our SAAS CereCloud. These Genres can encompass a number of different emotional styles, allowing CereWave Voices with these genres built in to seamlessly flow between them within the interface.
• LLM-powered word enhancement: CereProc’s engineers have leveraged the power of large language models (LLMs) to include word embeddings for all CereProc languages. This technology improves the natural intonation of the voice, with more accurate emphasis and speech rate. This means that you can have improved pronunciations of a words which will better match the context of the synthetic speech being generated by CereWave AI CereVoices.
• Faster Voice Cloning with CereVoice Me Version 2: CereProc can now generate synthetic voices using smaller amounts of data, with our exclusive AI based adapter models. CereProc’s AI based adapter models are unique, as they do not compromise the security of your voice data. CereProc, being an ethical voice cloning company, have built our AI based adapter models from only licensed and verified data.

The latest software update is now available on CereCloud, CereProc’s SAAS synthetic voice streaming service, which has users across the world. CereWave AI will soon be available to download through CereProc's official website and through the CerePlay App (Available on MacOS, iOS and Android). CereProc Users with software maintenance agreements are encouraged to update their CereProc software to take advantage of these enhancements and unlock the full potential of their CereVoices.

“CereVoice V7, incorporating CereWave AI V7 and an updated version of CereVoice Me – the worlds first voice cloning system at home, when launched in 2012, is yet another great step forward by Cereproc’ s engineering team headed by our CTO Chris Pidcock.” explains Paul Welham, CEO of CereProc, “CereVoice V7 not only offers higher quality synthetic speech on any platform, to meet the emerging needs and requests from customers, but also protects our users on older hardware devices while giving users new features to use. The latest version of CereVoice Me – continues our ethical stance of voice cloning, a practice not followed by many of our rivals, as recent events with fake voices has shown.”

For more information about the software update and other products and services offered by CereProc, please visit www.cereproc.com.

About CereProc Ltd.: CereProc is a Text to Speech company dedicated to developing innovative solutions that meet the requirements of our clients and customers. With a focus on innovative technology and accessibility, we strive to deliver exceptional experiences across all platforms.

For media inquiries or interview requests, please contact Willem Cluness at sales@cereproc.com