Malaysia’s AI for a Sovereign and Empowered Future
Empowering Malaysia through AI sovereignty by Agmo Group: data, hosting, and ownership in Malaysian hands.
The #1 Malaysian model, small enough to run in a school
MerdekaLLM-Sasbadi-27b, our education model fine-tuned with Sasbadi, scores 84.9% on MalayMMLU. At 27B parameters it can be hosted on-premise, so student data stays in the school.
27B
parameters, compact enough to host on-premise
84.9%
on MalayMMLU, the highest of the Malaysian-built models
1 GPU
to host it: a single 80 GB GPU holds the weights
- MerdekaLLM-Sasbadi-27b
- ILMU GLM-5.1
- Other models
Chart data and sources
| Model | Parameters | MalayMMLU | Size source |
|---|---|---|---|
| Kimi K3 | 2.8T | 87.5 | Model card |
| DeepSeek V4 Pro 0813 | 1.6T | 85.5 | Model card |
| Qwen 3.5 397B | 397B | 85.4 | Model card |
| MerdekaLLM-Sasbadi-27b | 27B | 84.9 | Agmo Group |
| ILMU GLM-5.1 | 744B | 84.0 | Model card |
| DeepSeek V4 Pro | 1.6T | 83.8 | Model card |
| MiniMax M3 | 428B | 82.8 | Model card |
| GLM-5.2 | 744B | 81.8 | Model card |
| Mistral Large 3 | 675B | 81.5 | Model card |
| Kimi K2.6 | 1T | 80.7 | Model card |
| Mistral Large 2512 | 675B | 80.4 | Model card |
| MiniMax M3 | 428B | 79.6 | Model card |
| Llama 4 Maverick | 400B | 78.1 | Model card |
| GLM-5.3 | 744B | 71.6 | Model card |
ILMU GLM-5.1 is fine-tuned from Z.ai’s GLM-5.1, so it has the same parameter count. Accuracy from the Pendakwah Teknologi leaderboard, except MerdekaLLM-Sasbadi-27b (internal evaluation). Total parameters; mixture-of-experts models use only part of them per token but must hold all of them in memory.
Source: Pendakwah Teknologi, 7 October 2026. Ours is an internal evaluation. GPU count is for the model weights alone.
Built by Malaysians, for Malaysia
Merdeka LLM is Malaysia’s AI Large Language Model designed to safeguard our digital future. Built entirely by Malaysians, hosted in Malaysian data centres, and trained on local data, it ensures that Malaysia’s voice is at the forefront of AI development, powered by Agmo Group.
LLM Training as a Service (TaaS)
In partnership with Phison’s aiDAPTIV+ and SNS, we offer scalable LLM training as a service, helping businesses unlock the power of AI for their own needs, from tailored language models to data management solutions.
LLM Gig Economy (Curatorship)
Malaysia’s first LLM Gig Economy Platform empowers Malaysians to participate in our AI’s growth through curated contributions, ensuring high-quality data inputs that keep Merdeka LLM relevant, accurate, and reliable.
AI Sovereignty as a Service
We empower enterprises and governments to achieve AI sovereignty with secure infrastructure, skilled talent, custom data solutions, and seamless deployment, keeping organisations in control of their AI systems and data.
Built here, for the way Malaysia works
Sovereignty Focused
Data privacy and security, with 100% Malaysian hosting and infrastructure.
Built for Malaysians
Tailored for Malaysian languages, cultures, and sectors while creating opportunities for Malaysians to contribute to AI development as data curators.
LLM Training as a Service
Leverage our partnerships with Phison’s aiDAPTIV+ and SNS to train your own AI models with the highest data security standards.
Where we’re headed
Mission
To lead Malaysia into the next digital age with a sovereign AI solution that empowers our industries, strengthens our data privacy, promotes innovation, and creates new opportunities for Malaysians through AI contributions and LLM training services.
Vision
A Malaysia where AI innovation thrives in full autonomy, protecting national interests while delivering cutting-edge solutions for businesses, government, and society with contributions from everyday Malaysians and strategic partnerships with leading AI and data infrastructure providers.
Let's Shape the Future Together
Join us in building a brighter tomorrow, reach out anytime.
Contact details
Level 38, MYEG Tower,
Empire City Damansara,
Jalan PJU 8, Damansara Perdana,
47820 Petaling Jaya, Selangor, Malaysia.