
Institute · Gothenburg, Sweden
AI Sweden
Sweden's national centre for applied AI, a public and private partnership that also published the GPT-SW3 Nordic language models.
Gothenburg, Sweden
- Founded
- 2019
- Type
- Institute
- Based in
- Gothenburg, Sweden
- Key papers
- 1
- Featured projects
- 4
About
AI Sweden describes itself as a national centre for applied AI. It is not-for-profit and works with more than 180 partners from companies, public bodies and universities.
Its programmes cover AI adoption and training, sector initiatives such as healthcare and automotive, a startup programme and research on infrastructure, secure AI and language technology.
In language technology it developed GPT-SW3 with RISE and WASP, a family of autoregressive models for Swedish and the other Nordic languages, with model cards dated December 2022 and an open release on Hugging Face reported in late 2023.
Mission and ambitionsAI Sweden states that its purpose is to accelerate the use of AI for the benefit of society.
Known for
- 01
GPT-SW3 Nordic language models
- 02
National applied AI partnership
- 03
Shared datasets and testbeds for Swedish AI
Milestones
- 2022
GPT-SW3 models documented
Hugging Face model cards for the first GPT-SW3 sizes are dated December 2022.
- 2023
GPT-SW3 paper
Technical description of the Nordic autoregressive models, posted to arXiv in May 2023.
- 2023
GPT-SW3 open release
Models from 126M to 40B parameters were made available on Hugging Face under a permissive licence.
Open source
Model: 1,470 downloads, 13 likes.
Model: 694 downloads, 12 likes.
- AI-Sweden-Models/BiaSWEdataset
Dataset: 134 downloads, 4 likes.
Model: 499 downloads, 7 likes.
Research papers
Notable releases
Openness
Publishes models and datasets on Hugging Face; individual licences vary, so check each model card.
People
- Magnus SahlgrenResearcher, GPT-SW3 co-author
- Ariel EkgrenLead author, GPT-SW3
- Joey ÖhmanResearcher, GPT-SW3 co-author
Listed from the lab's own pages and public profiles.
Contact and community
Contact through the website; staff addresses are listed individually so none is given here.
Sources
Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.



