Adyghe Goes Digital: Yandex Translate Adds One of Russia’s Indigenous Languages
Yandex Translate has officially launched text translation support for the Adyghe language, allowing users to translate to and from one hundred other languages worldwide. The feature is available both on the main website and in the mobile app.
The Adyghe-language translation system was developed through a large-scale linguistic data collection effort led by specialists from Adyghe State University, together with Yandex, the House of the Peoples of Russia and support from the Federal Agency for Ethnic Affairs.
A Symbol of Russia’s Cultural Diversity
Yandex Translate already supports 113 languages. Adding another rare language to the platform opens up major new opportunities for communication and for exploring the unique culture of the Circassian people.
“It is symbolic that the Adyghe language was added to the Yandex Translate platform during the Year of the Unity of the Peoples of Russia, declared by President Vladimir Putin. I congratulate everyone interested in Adygea, the regions of the North Caucasus and the Adyghe people as a whole who wants to learn more about the peoples of our large and diverse country,” said Murat Kumpilov, head of the Republic of Adygea.
Interest in the endangered languages of Russia’s Indigenous peoples is growing even as the number of native speakers declines and authentic speech becomes increasingly rare in everyday life. Now, however, these linguistic cultures are gaining a digital presence that could help preserve them for future generations.
By working with languages that have little existing digital data, Yandex is strengthening its position as a specialist in low-resource AI language systems. That creates a foundation for future features including voice input, speech synthesis, conversational translation and integration into search and educational platforms. Even today, users already have access to a practical digital tool for communication, learning and translation.
Endangered Languages Move Into the Spotlight
Major technology companies, universities and linguists are increasingly focusing on the rare languages spoken by Russia’s smaller Indigenous communities. In 2024, Yandex announced plans to add more than 20 previously unsupported languages of Russia’s peoples to Yandex Translate, with speech-recognition capabilities planned for some of them.
In April 2026, Sber reported that the updated GigaChat model had improved text understanding and generation across more than 30 languages spoken in Russia and CIS countries.
At Pyatigorsk State University, researchers developed an AI-based speech-recognition system capable of understanding languages spoken by Indigenous peoples of the North Caucasus. The technology could eventually be used in customer-support systems for government agencies and banks while also helping preserve endangered languages. The system has already learned to recognize Kabardian-Circassian and Balkar speech. That became possible after Kabardino-Balkarian State University provided developers with unique audio recordings from native speakers.
Back in 2022, Evenki language speaker Nikolai Aprosimo created Ayana, a mobile application translating Russian into the Evenki language, along with an online portal dedicated to learning Evenki.
Mission Possible
According to the Russian Academy of Sciences, 15 languages have disappeared in Russia since the beginning of the twentieth century. Another roughly 150 minority languages remain endangered. New technologies are becoming one of the tools used to slow that process. Digital archives of authentic texts, audio and video recordings are being created for each language, laying the groundwork for future translation and educational platforms.
Russia is now developing a distinct field centered on multilingual AI platforms designed to preserve cultural diversity. Existing systems are expected to expand by adding more rare languages, while IT engineers and software developers continue improving translation quality, voice input, speech synthesis and conversational modes.
The success of that effort will depend on close cooperation between researchers, developers and investors. One of the key conditions for long-term progress will remain the quality of linguistic data. The more actively universities, regional institutes, media organizations and native speakers contribute unique materials to digital archives, the more accurate neural-network services are likely to become.
