GoTo has released a 27 billion parameter large language model (LLM) trained in Bahasa Indonesia and four regional languages, making it freely available for startups, businesses, universities, researchers and public institutions to use, study and adapt (19/10).

The model was announced by PT GoTo Gojek Tokopedia Tbk at Bali Tech Week in Sanur, Bali, on 19 October 2026. Developed by the company's AI engineering team, it builds on GoTo's open-model work, which began in 2024.

GoTo Opens AI Model to Businesses and Researchers

GoTo said the model allows users to develop locally relevant AI applications without starting from scratch. Its release supports efforts to build AI technology that addresses Indonesia's local needs.

GoTo Group CEO Hans Patuwo said Indonesia needs to develop its own AI technology alongside using global AI tools.

"Indonesia makes great use of global AI, but we must build our own too, to solve the problems that are unique to us," Patuwo said.

He said AI developed for global use may not be fully optimized for local data or the regional languages that many Indonesians use at home. By making the model available to the public, GoTo aims to provide AI developers in Indonesia with a starting point based on local conditions.

The release also supports the Ministry of Communication and Digital Affairs' (Komdigi) Indonesia Digital Vision 2045. The initiative aims to create an inclusive, collaborative and innovation-driven digital ecosystem and help Indonesia move from being a consumer of AI technology to becoming a creator.

GoTo AI Model Trained in Indonesian and Regional Languages

The model was trained on Indonesian-language text from media partners and public-sector websites, alongside data from regional languages. It supports Bahasa Indonesia, Javanese, Sundanese, Balinese and Bataknese.

For Bataknese, GoTo worked with Universitas Sumatera Utara (USU) to collect data that reflects how the language is used today.

The model is designed to support language-related AI applications, including tools that answer questions, summarize documents and generate translations. Large language models learn language patterns from extensive collections of text, with the relevance and breadth of training data influencing their results.

27 Billion Parameter Model Designed for Wider Access

GoTo described the model as a mid-sized LLM with 27 billion parameters. Its size is intended to provide the capacity to handle demanding language tasks while allowing organizations to operate it on their own servers.

According to the company, universities, startups and public institutions can run the model without needing a large data center.

The model is freely available for organizations and individuals to use, study and adapt. This allows them to build on GoTo's existing technology rather than develop a language model from the beginning.

GoTo Develops Indonesia's Next Generation of AI Talent

GoTo is also working to develop the skills of people who will contribute to Indonesia's AI industry. Interns from seven universities, spanning Sumatra to Sulawesi and including Udayana University in Bali, have worked alongside the company's AI teams on practical projects.

Their work has included preparing data, training models and testing them. The interns also contributed their voices to the development of GoTo's Indonesian speech technology and helped collect Batak-language data.

GoTo's activities combine the development of locally relevant technology, open access to its model and opportunities for students to gain practical experience in AI development.

PHOTO: GOTO

This article was created with AI assistance.

We make every effort to ensure the accuracy of our content, some information may be incorrect or outdated. Please let us know of any corrections at [email protected].