Skip to main content

Regional AI integrates sovereign regional LLMs like Typhoon, HyperCLOVA X, and Rakuten AI into JITNA + SignedAI + DelentiaDB — delivering local compliance, regulatory context, and cultural nuance for Asia and global deployments.

Regional AI

Sovereign AI for Every Region

A pluggable AI OS infrastructure that allows developers to dynamically hot-swap national/sovereign LLMs to enforce local data laws, compliance, and cultural nuances.

🇹🇭

Standard Package — Typhoon v2

Active

Typhoon v2 70B by SCB10X is bundled as the default standard package for Thailand. However, the system is fully pluggable and lets developers register other native models at runtime.

G38 / Pluggable
Model ID
SCB10X
Provider
0.99
Proficiency TH
regional_core
Role

Supported Regional LLMs

Active standard packages are pre-installed. 'Pluggable' models can be registered in 3 simple steps.

🇹🇭
Thailand
Active
Typhoon v2 70B
SCB10X
Proficiency: 0.99
🇯🇵
Japan
Pluggable
Rakuten AI 3.0 / LLM-jp-4
Rakuten / NII
Proficiency: 0.98
🇰🇷
Korea
Pluggable
HyperCLOVA X / Solar Pro 2
NAVER / Upstage
Proficiency: 0.98
🇻🇳
Vietnam
Pluggable
ViGPT / Vistral
VinAI / Open Source
Proficiency: 0.95
🇮🇩
Indonesia
Pluggable
SEA-LION / Qwen-2.5-7B
AI Singapore
Proficiency: 0.95
🇮🇳
India
Pluggable
Krutrim / Sarvam AI
Ola / Sarvam
Proficiency: 0.96
🇸🇦
Saudi Arabia
Pluggable
ALLaM / Jais
SDAIA / G42
Proficiency: 0.96
🇨🇳
China
Active
Kimi K2.5 / MiniMax
Moonshot / MiniMax
Proficiency: 0.98

How to Plug in Your Regional LLM

The model registry supports hot-swapping and registering regional models dynamically at runtime.

1
Configure Model Entry

Define your model metadata, including context window size, input/output cost, and national compliance tags (e.g. APPI, PDPA).

2
Register Regional LLM

Call register_regional_llm(language, region, model_id, model_name, ...) dynamically at runtime.

3
Automatic JITNA Routing

JITNA automatically detects the user's language and region and routes calls to your custom LLM adapter.

# Register custom Japanese regional LLM dynamically at runtime
register_regional_llm(
    language="ja",
    region="JP",
    model_id="rakuten/Rakuten-AI-3.0-MoE-700B",
    model_name="Rakuten AI 3.0",
    proficiency=0.98,
    cost_input=0.40,
    cost_output=1.20,
    specialties=["Japanese legal", "Business documentation"],
    compliance_tags=["APPI"]
)

Why Typhoon + RCT = Complementary, Not Competing

🎯

Typhoon trains models

SCB10X trains LLMs specifically for Thai language and local context — producing models that understand Thai nuances better than general-purpose models.

🔗

RCT orchestrates models

RCT uses JITNA + SignedAI + DelentiaDB to route, verify, and persist results from Typhoon — putting Thai-language AI intelligence inside enterprise-grade infrastructure.

Ready to Deploy Your Regional AI?

Related Resources