# large language models
Latest news and articles about large language models
Total: 11 articles found
Two Spring Festivals, One Industry: How China’s Tech Giants Turned AI into a Holiday Battle for National Reach
China’s AI competition has shifted from model development to a consumer battleground during two consecutive Lunar New Year campaigns. Alibaba, ByteDance, Tencent and Baidu used subsidies, embedded assistants and viral features to fight for national traffic, while smaller firms pursue agent‑style products that combine multiple models. The outcome will reshape who controls mass AI touchpoints in China, narrow the US–China model gap and raise barriers for smaller players unless they adopt alternative, interoperable strategies.
China’s Kimi Rockets to a $10–12bn Valuation After Two Rapid Funding Rounds Exceeding $1.2bn
Kimi, a Chinese AI startup also known as Yue Zhi Anmian, has raised over $700 million in a new funding round led by existing investors, bringing two consecutive financings to more than $1.2 billion and valuing the company at $10–12 billion. The deals signal strong investor appetite for large AI models in China and a scramble by tech giants to secure model supply, but commercialisation and regulatory risks remain significant.
China’s Zhipu Pushes Prices Up as GLM-5 Goes Global — A Turning Point for Domestic AI Commercialisation
Zhipu Technology raised prices for its GLM Coding Plan and launched GLM-5 overseas on February 12, citing surging developer demand and the need for heavier investment in compute and model optimisation. The increase — 30% or higher domestically and substantially larger on overseas API pricing — marks a shift in China’s AI industry from low‑price competition to value-based monetisation.
Microsoft Bets on Homegrown AI, Predicts Rapid Automation of White‑Collar Work
Microsoft is pivoting from heavy reliance on OpenAI to building its own leading large language models, mobilising vast compute and teams and planning roughly $140 billion in AI‑related capital spending. CEO Mustafa Suleyman warned many desk‑based white‑collar tasks could be automated within 12–18 months, a claim that underlines both the opportunity and disruption in Microsoft’s strategy to deploy professional‑grade general AI for enterprises.
A Night of Acceleration: Zhipu’s GLM‑5 Debuts as MiniMax and DeepSeek Race to Keep Up
Three leading Chinese AI firms unveiled near‑simultaneous upgrades that signal a shift from demo‑level coding assistants to production‑oriented, agentic systems. Zhipu launched GLM‑5 as an open‑source foundation for long‑horizon engineering tasks, while MiniMax and DeepSeek pushed product and context upgrades aimed at real‑world throughput and extended interactions.
DeepSeek's Quiet Leap: 1‑Million‑Token Context and May‑2025 Knowledge Cut Hint at a Next‑Gen Chinese LLM
DeepSeek has begun limited testing of a model that supports a 1 million token context window and uses training data up to May 2025, a significant expansion from its previous 128k limit. The change suggests material architectural or pipeline upgrades and signals intensified competition among Chinese AI providers to ship more capable, enterprise‑ready models.
Amazon Circles a Custom OpenAI Model to Supercharge Alexa as Talks of a Large Equity Deal Advance
Amazon is negotiating with OpenAI on a commercial agreement that could include a multibillion-dollar equity investment and the creation of bespoke OpenAI models for Amazon products such as Alexa. The arrangement would deepen technical collaboration but raise strategic and regulatory questions about control, vendor lock-in and market concentration in AI.
Hang Seng Edges Higher as Tech Stocks Slip and AI Plays Rally
Hong Kong’s Hang Seng closed marginally higher while the Hang Seng Tech Index fell, as AI-related ‘large-model’ stocks and gold miners rallied and major internet platforms like Tencent slid. The session illustrated a market split: speculative AI plays attracted flows even as policy-sensitive mega-cap tech names remained susceptible to rumor-driven volatility.
Yang Zhilin Steps Forward: Moon’s Dark Side Ships Kimi K2.5 to Buy Time Against DeepSeek
Moon’s Dark Side released Kimi K2.5 with founder Yang Zhilin personally presenting the incremental upgrade, signalling a strategic shift from parameter-led competition to engineering improvements focused on coding and agent orchestration. The release is a defensive, deliverable move to shore up market position ahead of an expected DeepSeek model launch and to buy time for a more substantive K3 upgrade.
Alibaba Debuts Qwen3‑Max‑Thinking, a Tool‑Enabled Inference Model Aiming to Rival GPT‑5.2
Alibaba has launched Qwen3‑Max‑Thinking, an inference model that combines adaptive tool calling and test‑time scaling to improve reasoning, factual accuracy and alignment. Alibaba claims benchmark parity with leading models such as GPT‑5.2‑Thinking, and has deployed the capability in Qwen Chat, signalling rapid commercialisation within its cloud and consumer ecosystem.
China’s Kimi Says Algorithmic Ingenuity, Not Massive Compute, Powered Its Leap — and the AI Race May Be Changing
At Davos, Kimi’s leadership said it achieved state‑of‑the‑art results with its K2 series while using a fraction of the compute typical of leading US labs, crediting deep algorithmic and engineering innovation. The company plans to use fresh capital to expand hardware for a next‑generation K3 model, underscoring a broader Chinese push to compete via efficiency rather than brute‑force compute.