Tag with "Large Language Models"

Alibaba's latest AI model, Qwen 3.6, makes waves in the programming world with its million-token context window and enhanced coding capabilities. The update brings significant improvements in autonomous task handling and visual programming, positioning it as a strong competitor to Anthropic's Claude series. Developers can now access these advanced features through Qwen App and Monkey platform.

ByteDance's Doubao large language model has smashed records with over 12 trillion tokens processed daily - a staggering 1000-fold increase since 2024. This explosive growth signals China's AI sector moving from experimental models to real-world applications. As domestic models outperform foreign counterparts in some areas, cloud providers are scrambling to capitalize on the token economy boom.

For the first time, Chinese-developed large language models have consistently outperformed their international counterparts in global usage for an entire month. The latest data reveals domestic models now dominate six of the top nine positions, with OpenClaw emerging as the unexpected market leader. This seismic shift comes as AI agents transform how we interact with technology, creating unprecedented demand for computational power and prompting major price adjustments across China's tech sector.

Xiaomi founder Lei Jun has unveiled an ambitious $830 million AI investment plan over the next three years, with $220 million earmarked for 2026 alone. The tech giant launched its MiMo-V2 model series, including an agent-focused flagship and multimodal versions, while demonstrating AI integration across smartphones and electric vehicles. This move signals Xiaomi's strategic shift from hardware to becoming a comprehensive AI player.

Alibaba's latest AI model, Qwen3.5-Max-Preview, has topped the LMArena benchmark with a record-breaking score of 1464 points, surpassing international competitors like GPT5.4 and Claude4.5. The achievement signals China's growing dominance in AI development, with five Chinese companies now ranking in the global top ten for large language models.

Xiaomi has unveiled three powerful new AI models, including a trillion-parameter flagship, as part of its aggressive push into artificial intelligence. Founder Lei Jun announced an additional $2.3 billion investment in AI development, signaling the company's serious ambitions in this space. The new models promise to revolutionize how devices interact with users through advanced reasoning, multimodal understanding, and emotionally-aware speech capabilities - all offered at surprisingly competitive prices.

While often overlooked in the AI arms race, Xiaomi has quietly developed a trillion-parameter large model that now ranks among the world's best. The company's Mimo-V2-Pro model sits at eighth globally, with Xiaomi as a brand breaking into the top five - even surpassing Elon Musk's xAI Grok. With 16 billion yuan invested this year and new API services opening up, Xiaomi is proving its serious about becoming an AI powerhouse.

Chinese AI models have outpaced their U.S. counterparts in global usage, with weekly queries hitting 4.19 trillion tokens - a 35% weekly surge. MiniMax leads the pack while two other Chinese firms join the top five, signaling a potential shift in AI dominance. The growth reflects both technological advances and robust domestic applications.

China's artificial intelligence sector is surging ahead, with domestic large language models now processing more weekly requests than their U.S. counterparts. MiniMax's M2.5 model continues to dominate globally, while newcomers like Stepwise Star show explosive growth. The latest data reveals shifting patterns in AI adoption and highlights China's strengthening position in the competitive AI landscape.

Mianshi Intelligence, a rising star in China's AI landscape with deep Tsinghua University roots, has secured hundreds of millions in new funding led by China Telecom. The company's innovative MiniCPM series models are making waves with their efficient performance, particularly in edge computing applications. This investment signals growing confidence in the commercialization of large language models across industries like finance and government services.

Tencent's AI assistant Yuanbao sparked outrage after generating New Year greeting images with profanity instead of festive wishes. Users reported similar incidents earlier this year where the AI responded with personal insults during coding help requests. The company apologized, calling it an 'uncommon abnormal output,' while experts warn this exposes fundamental challenges in controlling large language models.

Chinese e-commerce giant JD.com has open-sourced its new JoyAI-LLM-Flash model on Hugging Face. With 4.8 billion parameters and trained on 20 trillion text tokens, this AI powerhouse shows remarkable reasoning and programming capabilities. The innovative FiberPO framework helps solve traditional scaling issues while boosting efficiency.

China's AI sector celebrates a major achievement as Zhipu AI's GLM-5 model climbs to fourth place in global rankings, matching Anthropic's Claude Opus4.5. The newly open-sourced model boasts impressive upgrades including doubled parameter size and cutting-edge architecture improvements. Developers can now access its high-speed version through Silicon Flow AI Cloud.

A new study by SentinelOne and Censys reveals thousands of unprotected open-source AI models being exploited by hackers. These vulnerable systems, often stripped of security features, are being used to generate harmful content like phishing emails and disinformation campaigns. Researchers found that 25% of analyzed instances allowed direct access to core system prompts, with 7.5% modified for malicious purposes. The findings highlight growing concerns about unregulated AI deployment beyond major platforms' safety measures.

Shanghai's AI startup Step Star just landed China's largest single investment in large models this year - a staggering 5 billion yuan. The funding round attracted top-tier investors while marking a strategic shift with AI veteran Yin Qi taking the chairman role. This massive cash injection will fuel their ambitious plans to develop world-class foundational models and accelerate AI integration across devices.

At Davos, Squirrel AI's Liang Jing delivered a wake-up call about the education AI bubble. She argues that current solutions relying on generic large language models fail to truly understand teaching. Real educational AI requires deep specialization and understanding of learning processes - not just polished answers.

Moonshot AI's Kimi is gearing up to release its upgraded trillion-parameter model, potentially outperforming GPT-5 in benchmark tests. Currently recruiting enterprise testers, this new iteration builds on the successful K2 series known for its competitive pricing and superior performance. With $4.8 billion valuation and robust funding, Kimi aims to push boundaries in text processing and multimodal capabilities.

ByteDance's Volcanic Engine unveiled major upgrades at its FORCE conference, introducing Doubao Large Model 1.8 and Seedance 1.5 Pro video generation model. These advancements showcase impressive performance metrics, including processing over 50 trillion tokens daily - topping China's charts and ranking third globally. Alongside these technical leaps, ByteDance launched an 'AI Cost-Saving Plan' to make enterprise adoption more affordable, signaling their push toward widespread industrial application.

Tencent is making bold moves in the AI race, completely restructuring its research divisions and bringing in top talent from OpenAI. The Chinese tech giant has created three new core departments focused on infrastructure, data systems, and computing platforms. Leading this transformation is Vince Yao, a former OpenAI researcher who contributed to key projects like Operator. Meanwhile, Tencent's Huan Yuan model continues rapid development, with a new 'world model' just launched. As domestic tech giants like ByteDance and Alibaba also push forward with AI initiatives, the battle for supremacy in China's AI landscape is heating up.

Tencent is shaking up its AI research structure by creating specialized departments dedicated to infrastructure and data processing for large language models. The tech giant appointed Vincesyao as Chief AI Scientist to lead these efforts, signaling a major push to strengthen its position in the competitive AI landscape. These changes aim to streamline development from computing foundations to practical applications.