Welsh language technology and artificial intelligence (AI): summary 2025 to 2026

Overview

Over the past few years, we've been working hard to improve technology to help more people to use Welsh in everyday life.

Here are our current priorities:

• improving technology to increase the daily use of Cymraeg
• making sure everyone can access Welsh language technology
• improving Welsh language artificial intelligence (AI) and speech and language technologies (by sharing data and other means)

Here's a summary of our work for 2025 to 2026.

The latest is available at Welsh language technology and artificial intelligence (AI).

A summary of what we achieved last year is available at Welsh language technology and artificial intelligence (AI): summary 2024 to 2025.

Geraint and Gwyneth's Welsh synthetic voices are now available through Hwb

We want to make sure everyone can access Welsh language technology. An important part this is helping people with specific needs.

Through a collaboration between the Welsh Government and the Royal National Institute of Blind People (RNIB), learners and teachers can now download the voices directly from Hwb and use them with compatible assistive technology.

With funding from the Welsh Government, the RNIB developed these voices with software company IVONA. The voices are free for individuals and organisations to use for non-commercial purposes. You can now install them on school or home devices and use them with any Windows assistive package, including JAWS, NVDA, Supernova, Zoomtext, Read and Write, Clicker 8 and Tobii Dynavox.

This work isn't only so that people can use more of their Welsh but also so that they can use it easily, without additional challenges.

Download the voices from Hwb.

Microsoft Partnership

Since we worked with Microsoft to createa simultaneous translation facility in Teams, and since Microsoft announced that they now support Welsh and Catalan in Copilot, Microsoft Teams now has the ability to transcribe Welsh and English at the same time. We continue to work in partnership with Microsoft to support bilingual working here in Wales and beyond.

Cymraeg mewn Clic

We published 'Cymraeg mewn clic', a one-click app on our corporate network that reduces "friction" (behaviour change methodology) in the use of Welsh in technology; Cymraeg mewn Clic makes it easier for more of our staff to have a full experience of the Welsh language on their work laptops.

Technology grants

The technology grants continued to help make more Welsh available in more spaces. From AI models and new research resources to thousands of Wikipedia articles and Welsh-language geographic data that are used every day, the projects have helped build the digital infrastructure that the Welsh language needs to thrive in the digital age. Together, they have created new content, shared open data, developed useful tools and put the Welsh language at the heart of new technology. The work also contributes directly to our Welsh language technology and AI priorities, strengthening the resources, data and research capacity needed to develop future Welsh language technology.

Here are some highlights:

Iriaith (Bangor University)

• The project collected nearly half a million new English-Welsh sentences for Welsh-language technology research and development. This open data now helps to train AI models and translate Welsh (Llyw Cymru dataset).
• The team developed new Welsh language AI models, including machine translation models and large language models that can respond in Welsh (Welsh Llama Model).
• The project improved Welsh speech technologies, developing more accurate transcription models and new synthetic voices that sound more natural (Transcriber and TTS Cymru).
• Useful services such as Macsen and Trawsgrifiwr have been developed which put that research into practice and help more people to use the Welsh language in everyday life and work (Macsen).

Want to know more about the work of the Language Technologies Unit at Bangor University? Here is a link to the Welsh National Technologies Portal.

Mapio Cymru (Data Orchard)

Welsh geographic data helps to make the Welsh language more visible in the digital world. Mapio Cymru has expanded that data and ensured that it is used by more organisations, services and users across Wales.

• The project increased the amount of Welsh geographic data available openly, adding over 30,000 new items to OpenStreetMap.
• The project's website reached over a million unique visits during the year, bringing Welsh place names to new audiences (OpenStreetMap Cymru).
• The project's data has been used by major organisations such as the Senedd and Estyn, helping to make the Welsh language more visible in everyday digital services.

Wici-Fflics (National Library of Wales)

Welsh content online is one of the most important resources we have to support the use of the Welsh language today and the development of Welsh technologies tomorrow. Wici-Flics helped expand that content on a significant scale by combining open data, AI, and human editing.

• The project created over 1,000 new Welsh Wicipedia articles about people in film and television (Wici-Flics Project).
• Over 7,000 people, places and organisations have been added to open bilingual databases, strengthening the resources available to the public and developers (SNARC Fund).
• The project used AI in a responsible way to help increase the amount of information available online, while retaining human oversight over the content.
• The public has also been working and experimenting with this data in a History Hackathon. Here's a blogabout the event.

CEFR Analyser (Cardiff University)

This project helped build the resources and expertise needed to support the future work of CEFR (Common European Framework of Reference for Languages).

• The project developed a new CEFR-Cymraeg dataset to help technology understand and measure Welsh proficiency levels.
• It explored how modern AI models can support Welsh assessment and learning more effectively.
• Models were then openly shared with the research community (Welsh CEFR model collection).

Inquiry "Cymraeg i Bawb?"

The then Cabinet Secretary for Finance and the Welsh Language appeared before the Senedd's Culture, Communications, Welsh Language, Sport, and International Relations Committee on 16 July 2025. Technology and the Welsh language were part of the discussion "Cymraeg i bawb?". The evidence paper which shows the discussion in full, and a video of the session is available.