
Academic lab · Tartu, Estonia
TartuNLPUniversity of Tartu
TartuNLP is the University of Tartu's language technology group, building Estonian and Finno-Ugric models, translation and speech tools.
Tartu, Estonia
- Type
- Academic lab
- Based in
- Tartu, Estonia
- Key papers
- 2
- Featured projects
- 8
About
TartuNLP is a natural language processing group at the University of Tartu in Estonia. Its site presents work on Estonian language models, machine translation and speech synthesis.
Named projects include Llammas, an Estonian-adapted Llama model, EstBERT, and the Neurotõlge translation service. Code and models are published through the TartuNLP GitHub organization.
Mission and ambitionsThe group works to provide open language technology for Estonian and related smaller languages.
Known for
- 01
Estonian language models
- 02
Neurotõlge machine translation
- 03
Finno-Ugric language tools
Milestones
- 2020
EstBERT
An Estonian BERT model is published.
- 2023
Llammas
An Estonian instruction-tuned Llama model is released.
Open source
Estonian multi-speaker neural text-to-speech worker (20 stars)
Scripts for finetuning M2M-100 models (19 stars)
API for neural text-to-speech synthesis (17 stars)
REST API for neural machine translation (14 stars)
- tartuNLP/EstBERTmodel
Estonian BERT model.
70B Estonian instruction model.
Benchmark for Finno-Ugric languages.
- tartuNLP/XTTSv2-estspace
Estonian XTTS speech synthesis demo.
Research papers
Notable releases
Openness
Models and datasets are public on Hugging Face; code on GitHub is mostly MIT.
People
- Mark FišelAssociate professor, head of the group
Author of the EstLLM report.
- Kairit SirtsAssociate professor
Co-author of EstBERT and EstLLM.
- Tanel AlumäeResearcher (also TalTech)
Speech technology co-author on EstLLM.
- Taido PurasonResearcher
Co-author of EstLLM and Llammas-related work.
Listed from the lab's own pages and public profiles.
Contact and community
The group's website and GitHub organisation are the public entry points; its site could not be read as plain text.
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 Something wrong or out of date?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.



