An emotional speech synthesis model in the Kyrgyz language has been presented in Kyrgyzstan. The model was developed by the teams behind Nineninesix and Akylai projects and has been released as open-source for testing and integration into technological solutions across the public and private sectors. The press service of High Technology Park reported.
The model is designed to generate Kyrgyz speech capable of conveying intonation and emotional nuances. The release marks a significant milestone in the digital development of the Kyrgyz language, laying the foundation for the creation of localized voice services, virtual assistants, and educational tools.
The public release allows developers, companies, and research groups to:
- use an open-source Kyrgyz-language voice engine in their products and prototypes;
- implement voice assistants and call center solutions in Kyrgyz;
- create adaptive learning systems and tools for audio-based education;
- improve digital accessibility for users with special needs;
- localize and voice digital content in the Kyrgyz language.
This initiative helps build a full technological stack for Kyrgyz speech processing and expands the language’s presence in digital products.
The model generates speech in real time and can operate on standard GPUs commonly found in consumer computers.
«We believe that an open ecosystem around Kani TTS will accelerate the emergence of new Kyrgyz-language services — from voice interfaces in the public sector to localized business solutions,» the developers said.
