Asian AI
AI models, research labs, and language technology built in Asia.
19 articles
DeepSeek V4 Is Missing In Action, And China's AI Pride Is Sweating
DeepSeek's long-rumoured V4 is still not in the API. Hangzhou's most-watched lab now has a credibility problem.
Taiwan's TAIDE Is Powering a Wave of Traditional Chinese Creators
Sovereign Taiwanese AI is winning where global models stumble: cultural register.
Seoul's Silent AI Revolution: How HyperCLOVA X Think Became Part of Korean Daily Life
Korea's sovereign LLM has quietly become the default assistant for an entire country.
Sarvam AI's Record India Series C: Bengaluru Finally Has a Real Sovereign AI Champion
One of India's largest AI raises hands Sarvam the sovereign LLM seat at India's top table.
DeepSeek AI: Free GPT-5 Rivals Just Arrived!
DeepSeek's open-source models challenge American AI dominance, offering comparable performance at 70% lower cost. A strategic inflection point for Asia-Pacific development.
Yi-Lightning From 01.AI: How Kai-Fu Lee's Model Is Quietly Winning Enterprise Deals
01.AI's Yi-Lightning model has attracted less attention than DeepSeek but is building a stronger enterprise book of business across China's financial and insurance sectors.
The Singapore AI Model Ecosystem: What NUS, A*STAR and GovTech Are Building
Singapore's strategy is not to compete with frontier models but to build specialised AI that the US and China cannot or will not build. The results are starting to show.
InternVL3: How Shanghai AI Lab's Multimodal Model Became Asia's Vision AI Benchmark
Shanghai AI Laboratory's InternVL3 has set new standards for open-source vision-language models and enabled a generation of Asia-specific derivatives.
Doubao 1.5 Pro: How ByteDance Built China's Most Used AI and What It Means for the Region
ByteDance's Doubao has quietly become the most used AI assistant in China with 60 million daily active users. The 1.5 Pro update raises the ceiling — and the regional stakes.
MiniMax M2.7: The $0.30 Chinese Model That Evolves Itself
China's MiniMax drops a self-evolving AI that rewrites its own code and costs 50x less than Western rivals.
Wan 2.1: Why China's Open-Source Video AI Has Changed the Creative Industry
Alibaba's Wan 2.1 has become the reference video generation model for AI-assisted production across Asia. Here is how it happened and what comes next.
HyperCLOVA X Think Is Korea's Reasoning Model — And It Is Better Than Expected
Naver's reasoning-tuned variant of HyperCLOVA X demonstrates that Korean AI labs can compete with international models on the reasoning benchmarks that matter most for enterprise use.
NTT and SoftBank Are Building Asia's First Frontier-Scale Sovereign Model
Japan's two largest telecommunications companies have formed an unprecedented AI partnership targeting a 100-billion-parameter model. If it succeeds, Japan finally controls its own AI frontier.
Krutrim Is Building Something That Does Not Exist Yet: An AI for 22 Indian Languages
Ola's AI lab Krutrim has released its second generation model with dramatically improved performance across India's linguistic diversity. Here is why that matters beyond India.
Qwen 3 Arrives: Alibaba's Model Family Just Became Impossible to Ignore
Alibaba's third generation of the Qwen series raises the bar for open-source LLMs and lands a credible challenge to both Western frontier models and ByteDance's Doubao.
SeaLION 3 Is Here: Why Southeast Asia's Own LLM Is Finally Ready for Production
After two years of development, AI Singapore's SeaLION model family has reached a level of capability that makes it a credible option for enterprise deployment across the region.
Tencent Takes on DeepSeek: Meet the Lightning-Fast Hunyuan Turbo S
Tencent launches Hunyuan Turbo S with sub-second response times, directly challenging DeepSeek's AI dominance in China's heated competition.
DeepSeek Dilemma: AI Ambitions Collide with South Korean Privacy Safeguards
South Korea blocks DeepSeek downloads over ByteDance data sharing, setting precedent for AI regulation across Asia's privacy-conscious markets.
DeepSeek's Rise: The $6M AI Disrupting Silicon Valley's Billion-Dollar Game
DeepSeek, a two-year-old Chinese AI startup, reaches #1 on US App Store with just $6M in training costs, challenging Silicon Valley giants who spend $100M+.