Pathumma LLM

Thailand's open multimodal government LLM from NECTEC, handling Thai text, vision and audio.

LLMs & Chat Open Source Has API Open Source
Researched · Published · Reviewed
RECATOOLS Score
6.5 / 10
HQ
Thailand
Launched
2024
Developer
NECTEC (NSTDA)

Overview

Pathumma LLM is an open multimodal large language model developed by NECTEC under Thailand's NSTDA, tailored to Thai language, context and culture. It combines three core capabilities: a text LLM for Thai, a vision LLM for image analysis, and an audio LLM for Thai speech. Built with a focus on official and public-sector documents, it has been adopted by the Thai Parliament for tasks like summarizing meetings and legislative drafts. It is offered as a web app, an API for developers, and downloadable open models.

Advertisement

Pricing

Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Sep 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.

Free
Free
Free web app plus open downloadable models and a developer API.

What you can produce with Pathumma LLM

  • Run an open Thai model that handles text, vision and audio rather than text alone
  • Work in Thai language and cultural context rather than translating through English
  • Self-host it inside Thailand, which is the point of a government research model
  • Use the vision component on Thai documents and the audio component on Thai speech
  • Build on NECTEC's work under NSTDA, Thailand's national science agency
  • Judge it on research standing — there is no commercial review presence to consult
Advertisement

ASEAN Perspective

Pathumma LLM in Southeast Asia

Pathumma is a multimodal model from NECTEC that processes Thai speech as well as text and images. This matters in a country where much official business is conducted orally and where general-purpose models struggle with Thai audio. Unlike a pilot project, it's in active use: the Thai Parliament uses it to summarise meetings and legislative drafts. It's a rare public example of a government research institute successfully shipping a multimodal model tuned to a national language.

RECATOOLS Verdict

What this is for: A multimodal Thai LLM for text, image and speech, tuned for government and public-sector documents.

Who this is for: Thai government agencies, developers and researchers needing localized multimodal AI for official Thai-language work.

Availability: Open and free via a web app, developer API and downloadable models on Hugging Face.

Independent AI-assisted assessment by RECATOOLS.

What people say

As an open government research model from NECTEC (NSTDA), Pathumma LLM doesn't have a presence on commercial review sites. It's an open government research model from NECTEC (NSTDA), so the only signals that matter are institutional adoption and coverage in the Thai press. The most concrete sign of uptake is its use in the Thai Parliament for summarizing meetings and legislative drafts—a real-world deployment, not just a press release. National outlets (Nation Thailand, Thai Times) and NSTDA covered it as a flagship localized Thai model.

It's a family of separate models for text, vision, and audio (including an 8B system for Thai speech), all published on Hugging Face. The text model was benchmarked on Thai-specific suites like m3exam and thaiexam. An open request to add NECTEC's models to Stanford's HELM shows some outside interest in independent evaluation.

The model's public record also reveals its limitations. Independent head-to-head benchmarks are thin. In Thailand's crowded local-LLM scene, it competes with well-regarded models like SCB 10X's Typhoon and OpenThaiGPT, and doesn't clearly lead on published numbers. The system is split across multiple checkpoints that you have to integrate yourself. And with most coverage originating from NECTEC/NSTDA itself, neutral evaluation is scarce.

Summary of public user & expert reviews, compiled by RECATOOLS.

Frequently asked questions

It is multimodal where the others are text-only: a Thai text model, a vision model for image analysis and an audio model for Thai speech.
Yes. The Thai Parliament has adopted it for work such as summarising meetings and legislative drafts.
Three ways — a web app, a developer API, and downloadable open weights on Hugging Face.

About this listing

Researched on
Published on
Last reviewed

This entry was compiled from publicly available data including Pathumma LLM's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Pathumma LLM unless explicitly stated.

Data accuracy

Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.

For the latest details, please refer to Pathumma LLM directly →

Spotted something out of date? Suggest an update →

Advertisement