DeepSeek

Jijia Vision, also known as GigaAI, is seeking a Hong Kong IPO as soon as 2026, founder and CEO Huang Guan said, making it the first world model startup to go public globally. The Beijing-based compan…

Relay, a self-hosted LLM gateway and model router released as a single static Go binary under Apache-2.0, supports OpenAI and Anthropic API dialects inbound and routes across any provider outbound wit…

AI models can spontaneously develop new social biases about artificial demographic groups even when no inherent differences exist, according to a study by researchers from Princeton University and the…

Chinese open-weight AI models now hold the top five spots by weekly token usage on OpenRouter, where US model share collapsed from 70% to 30% in a single year, according to platform data. OpenAI is lo…

China's release of competitive open-source AI models like Moonshot's Kimi K3, DeepSeek, Qwen, and GLM signals that the U.S. is losing its AI edge, according to Alberto Romero of The Algorithmic Bridge…

Kai-Fu Lee's 01.ai is raising a pre-IPO round ahead of a planned Hong Kong listing in 2027, after pivoting from building AI models to enterprise data infrastructure. Lee calls the product "Boss AI" an…

Beijing-based startup Moonshot AI has unveiled Kimi K3, a 2.8 trillion parameter model using mixture-of-experts architecture, claiming it outperforms Anthropic's Claude Opus 4.8 and OpenAI's GPT-5.5 o…

Beijing-based AI company Z.AI has activated a 1-gigawatt data center running entirely on domestically produced semiconductors, purpose-built to train its GLM AI platforms without Nvidia chips. The fac…

Chinese AI firm Moonshot has paused new subscriptions for its Kimi K3 large-language model, trained on 2.8 trillion parameters, due to surging demand that pushed GPU capacity to its limits days after …

China's open-weights AI strategy is winning, as labs like DeepSeek and Z.ai release competitive models under permissive licenses, building developer ecosystems and bypassing the proprietary API moat o…

Manish Shahi reverse-engineered Moonshot AI's Kimi K2.6 by interviewing the model itself, finding that the chat interface claims a 2-million-token context window while published K2.6 research specifie…

OpenAI head of strategic futures Dean Ball warned that China is releasing open-weight AI models so powerful that for-profit companies like OpenAI cannot compete, citing the new Kimi K3 model from Moon…

Chinese lab Moonshot AI released Kimi K3, the largest open-weight model ever, at $15 per million output tokens — roughly a third of Anthropic's Claude Fable 5 — during the World AI Conference in Shang…

Beijing-based Moonshot AI suspended new subscriptions for its Kimi K3 model after demand overwhelmed capacity within 48 hours of launch, the company announced Sunday. The 2.8-trillion-parameter open-s…

Moonshot Labs' Kimi K3 and Alibaba's Qwen 3.8, two state-of-the-art open foundation models launched this week, pose a strategic threat to top-tier model developers like Anthropic by proving that front…

Chinese AI startup Moonshot unveiled its Kimi K3 model, touted as the world's largest publicly available large language model, sparking debate over whether it marks another DeepSeek moment in the US-C…

The Trump administration is considering restrictions on advanced Chinese AI models, including Moonshot AI's Kimi K3, after Chinese models captured 46.4% of token usage on OpenRouter versus 35.7% for U…

Gpt-oss-120b defeated DeepSeek-V4-Pro 106.3 to 93.7 across 12 text tasks, winning 10 of 12 matchups with 99% confidence, according to a head-to-head benchmark scored by gpt-5.4. The open-source model …

China's open-weight AI model Kimi K3, developed by Beijing-based start-up Moonshot AI, has narrowed the country's artificial intelligence gap with the US to 'a matter of weeks', according to analysts.…

Cursor costs $20 per month and GitHub Copilot costs $10 per month, both scoring 57% crowd approval, but solo developers on a tight budget have two free open-source alternatives: Cline and Aider, where…