
Academic lab · Rayong, Thailand
VISTEC-depa AI Research Institute of Thailand
VISTEC-depa AIRI is a Thai institute publishing Thai NLP datasets and models such as WangchanBERTa and WangchanLION.
Rayong, Thailand
- Type
- Academic lab
- Based in
- Rayong, Thailand
- Key papers
- 1
- Featured projects
- 8
About
The institute was formed through a collaboration between depa and VISTEC. It provides large datasets and pretrained models for Thai.
WangchanBERTa is its Thai transformer language model. WangchanLION, built with AI Singapore on SEA-LION, is released with code, data and weights under Apache 2.0, and WangchanLION-v3 continues pretraining from Llama 3.1 8B.
Mission and ambitionsThe institute aims to provide Thai language AI resources.
Known for
- 01
WangchanBERTa
- 02
WangchanLION
- 03
Thai datasets
Milestones
- 2020
English-Thai translation dataset and models
Released in June.
- 2021
WangchanBERTa
Thai RoBERTa-based model pretrained on 78 GB of text, released in March.
- 2021
Thai Speech Emotion Dataset
Released in March.
- 2024
WangchanLION
Instruction-tuned Thai SEA-LION model.
- 2025
WangchanLION-v3
Continual pretraining from Llama 3.1 8B.
Open source
Pretraining Thai transformers.
Thai wav2vec2 fine-tuning.
Kaldi recipe for Thai Common Voice.
Dataset releases.
Thai ASR.
WangchanBERTa.
Thai instruction data.
- airesearch/thai-serdataset
Thai speech emotion data.
Research papers
Notable releases
Openness
Models and datasets on Hugging Face.
People
- Sarana NutanongResearcher at VISTEC, WangchanBERTa co-author
Author list of the WangchanBERTa paper.
Listed from the lab's own pages and public profiles.
Contact and community
airesearch.in.th lists the institute's contacts.
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 Something wrong or out of date?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.



