Why Azendo About Azendo Our values Talent Success & Academy Why Thailand Offices and facilities
Services Dedicated software development teamDedicated mobile development teamDedicated UI/UX design teamDedicated QA and testing teamDedicated cloud and DevOps teamDedicated data engineering teamDedicated AI engineering teamDedicated AI automation teamDedicated MLOps teamDedicated data science team
Industries SaaS and technology All industries
How it works Pricing Newsroom Career Get in touch

AI in Thailand: the people, models and investment behind Southeast Asia's quiet tech riser.

In four years Thailand has gone from publishing an AI strategy to putting more than a million people through AI training, releasing several homegrown Thai language models and bringing the world's largest cloud providers onshore. The story of AI in Thailand is less about data centres than about the engineers coming through behind them.

By

Illustration of AI in Thailand: two people working at laptops either side of an arc, a sun between them
Thailand's AI story is being written in training rooms as much as in server halls.

Thailand’s AI strategy came with deadlines

The story of AI in Thailand has one feature that sets it apart from many national AI plans: it comes with numbers attached. The Cabinet approved the National AI Strategy and Action Plan on 26 July 2022, covering the years 2022 to 2027. In 2025 a National AI Committee chaired by the Prime Minister went further and set targets for the next two years: ten million Thais with working AI literacy, 90,000 AI professionals and 50,000 AI developers.

Targets on paper are easy. What makes the Thai case worth following is that the training programmes behind them are reporting results, and some of those results are ahead of schedule.

Thailand's AI workforce in four numbers

1.58mThais trained in AI through THAI Academy in Microsoft's 2025 financial year, against a target of one million
45,800+registrations for the Super AI Engineer programme across five seasons
1,224engineers who completed the demanding final level in season five alone
90,000AI professionals the national committee wants in place within two years
Sources: Microsoft, the Artificial Intelligence Association of Thailand and the Bangkok Post. The first three are reported results, the fourth is a target.

Thai AI training at a scale few countries attempt

The largest of these programmes is THAI Academy, launched in April 2025 by the Thai government with Microsoft. Its public goal was to give one million Thais AI skills by the end of 2025. Microsoft reports that the programme reached 1.577 million people in its 2025 financial year, well past that target, including more than 100,000 civil servants and over 20,000 small and medium sized businesses.

AI skills training: target against reported result

Public target, by end of 20251.00 million
Reported, Microsoft FY20251.577 million
The next phase aims at 250,000 teachers, 100,000 workers and 77 training centres, on the way to ten million Thais by 2027. Source: Microsoft.

Broad literacy is only the base of the pyramid. Building AI products needs engineers, and here the most interesting institution is not a ministry or a multinational but a professional association.

Super AI Engineer: a national filter for serious engineers

The Artificial Intelligence Association of Thailand runs Super AI Engineer, a free programme funded through the country’s public innovation budget. Anyone can register. Very few make it to the end. Candidates work through three levels of training, competitions and team projects, and each level cuts the field hard.

Super AI Engineer, season five

Registered12,865
Reached level 21,607
Completed level 31,224
Fewer than one registrant in ten completes the final level. Source: Artificial Intelligence Association of Thailand.

Across its first five seasons the programme has drawn more than 45,800 registrations, and season six opened on 13 March 2026. Because the selection is so steep, the certificate means something. It tells an employer the holder has done sustained, hands on work, not a weekend course.

The association also works with industry. A 2026 AI engineering programme it ran with Microsoft between February and May certified 444 participants, who put forward 433 project concepts. Ninety three were selected for development and 53 teams reached the final. The concepts were practical: public services, agriculture, healthcare and finance, built by people who know how those systems work in Thailand.

Thai AI models built for the Thai language

Thai is a hard language for AI. It has its own script, five tones, and it is written without spaces between words, so even working out where one word ends and the next begins is a problem a model has to solve. Global models handle it, but not as well as they handle English. Thailand’s response has been to build its own.

Homegrown Thai language models

SCB 10XTyphoonOpen Thai models released from late 2023. The second generation spans small and large text models plus vision, audio and safety models.
NECTECPathummaThe national electronics and computer technology centre's Thai model family, released in 2024.
AI Association of ThailandOpenThaiGPTA community driven open model effort built with academic and industry partners.
KBTGTHaLLEA finance focused model from Kasikorn Bank's technology arm, tested on professional finance exam material.
Big Data InstituteThaiLLMA national project coordinating Thai model development across research institutes and companies.
Several of these are published with open weights, so Thai companies can run them on their own infrastructure.

The significance goes beyond national pride. A bank, hospital or government agency that needs AI to read Thai documents, answer Thai customers or search Thai legal text now has local options, built by people who understand the language and its edge cases. It also means Thailand has engineers who have trained, evaluated and shipped large language models themselves, which is a rarer skill than using one.

AI in Thailand’s hospitals and banks

Applied AI in Thailand is already past the pilot stage in some sectors. A medical AI consortium led by NECTEC has developed more than ten models for clinical use, and at least two have been approved by Thailand’s Food and Drug Administration. The large Thai banks, through technology arms such as SCB 10X and KBTG, run research teams that publish papers and release models rather than simply buying software.

International technology companies have noticed the workforce too. Agoda, the travel platform, opened its largest office in Bangkok, a technology hub at One Bangkok built for nearly 4,000 people. A company that size does not build its biggest engineering base somewhere it cannot find engineers.

Thailand’s AI infrastructure is arriving too

Talent needs somewhere to run its work, and the large cloud providers have now committed to Thailand. AWS opened its Asia Pacific (Thailand) Region in January 2025 with a planned investment of more than US$5 billion. Google announced a US$1 billion investment in 2024 and its Bangkok cloud region has since gone live. Microsoft announced its first datacenter region in the country in 2024. Locally, Siam AI operates GPU infrastructure as an NVIDIA Cloud Partner. The Bangkok Post reports AI infrastructure investment commitments of more than US$15 billion in total.

For companies building software here, the practical effect is simple: data can stay in the country, latency is low, and engineers work on the same platforms as their counterparts in Europe.

Milestones, 2022 to 2026

  1. July 2022National AI Strategy approved, setting the plan through 2027.
  2. Late 2023SCB 10X releases Typhoon, one of the first open Thai language models.
  3. 2024NECTEC releases Pathumma. Microsoft and Google announce cloud investments in Thailand.
  4. January 2025AWS opens its Thailand region.
  5. April 2025THAI Academy launches with a target of one million people trained in AI.
  6. 2025National AI Committee sets two year targets, including 50,000 AI developers.
  7. By January 2026Google's Bangkok cloud region goes live. KBTG publishes its THaLLE work with the ThaiLLM project.
  8. March 2026Super AI Engineer season six opens.
  9. May 202653 teams reach the final of the Microsoft and AIAT AI engineering programme.
Four years from strategy to working models, trained engineers and local cloud regions.

Who is building Thailand’s AI ecosystem

The ecosystem at a glance

Policy and public researchGovernmentNational AI Committee, NSTDA and NECTEC, the Digital Economy Promotion Agency, and the Big Data Institute.
Training and communitySkills pipelineThe AI Association of Thailand and Super AI Engineer, THAI Academy, and the universities feeding both.
Thai companiesBuildersSCB 10X and Typhoon, KBTG and THaLLE, Siam AI's GPU cloud, and a growing number of software companies.
Global technology firmsInvestors and employersMicrosoft, Google and AWS on infrastructure and training, and Agoda among the large technology employers.
What stands out is how much the four groups work together. Most of the programmes above involve at least two of them.

Where AI in Thailand goes next

In the Global Innovation Index, Thailand ranks fourth among upper middle income economies and has been rated an innovation overperformer for its income level eleven times.

Business adoption is the next stage. Government and the large banks are leading it, and small and medium sized companies are starting now. That is why the training targets are so ambitious, and why so many newly trained engineers are looking for serious product work.

What AI in Thailand means if you build software

We have delivered software from Chiang Mai and Bangkok since 2010, and the change in the last three years is visible from the inside. Engineers now arrive with hands on experience of retrieval, evaluation and agent workflows, often from Super AI Engineer or a similar programme, and many of them have worked with Thai language models that were built a few kilometres away.

For a European company, that matters for three reasons. The skills now exist in depth rather than in a handful of specialists. The cloud platforms are local. And the working day overlaps well with Europe, which is part of why we deliver from Thailand in the first place.

As an offshore software development company, we assign engineers to one client roadmap for the long term, and every specialist goes through our own AI certification before assignment, described under Talent Success and Academy. If you are planning AI work, companies hire AI developers through us as a dedicated team for model-backed features in production, and our guide to AI automation tools explains how we build agent workflows. For a wider view of the market and who builds software here, read our piece on IT outsourcing Thailand, or see how a dedicated development team from Thailand is set up.

Sources

Questions about AI in Thailand.

Is Thailand a good place for AI development?

Increasingly, yes. The country has a national AI strategy running to 2027, free national training programmes that screen thousands of engineers every year, its own Thai language models from both public research institutes and private companies, and cloud regions from AWS, Google and Microsoft. Business adoption is still at an early stage, which is worth knowing, but the supply of engineers who can build with AI is growing quickly.

Which Thai language AI models exist?

The best known are Typhoon from SCB 10X, Pathumma from NECTEC, OpenThaiGPT from the Artificial Intelligence Association of Thailand, and THaLLE from KBTG, which focuses on finance. A national ThaiLLM project coordinated by the Big Data Institute brings several of these efforts together.

What is the Super AI Engineer programme?

A free national programme run by the Artificial Intelligence Association of Thailand with public innovation funding. Candidates move through three levels of increasingly demanding training and project work. It has drawn more than 45,800 registrations across its first five seasons, and the sixth season opened in March 2026.

How does Thailand rank on innovation?

Thailand placed 45th of 139 economies in the Global Innovation Index 2025, down from 41st the year before, and fourth among upper middle income economies. It has been rated an innovation overperformer for its income level eleven times.

Is AI widely used by Thai businesses yet?

Not yet. Surveys reported in the Thai press suggest fewer than one in five of the organisations asked had adopted AI by 2024. Government and large banks are ahead, and smaller companies are only starting. That gap is exactly what the national training targets are meant to close.

Tell us what the roadmap needs and we will tell you what it takes.

A service delivery manager replies with the disciplines we would assign, the monthly capacity and what the first month looks like.

Get in touch How it works

Thinking about delivering from Thailand.

Loading the contact form… You can also email hello@azendo.co.

We reply within one working day. No obligation, and no newsletter.