Tag
Ben Thompson proposes US legislation to explicitly allow training data as fair use and bar terms of service that forbid distillation, arguing this could help US open models compete with Chinese counterparts, and suggests Alibaba's release of Qwen 3.8 Max may have been influenced by Xi Jinping's call for open source collaboration.
OpenAI's Head of Strategic Futures discusses the implications of open-weight Chinese AI models.
Chinese AI models are 112x cheaper than Anthropic per million tokens, reflecting the steepest commodity curve in technology history, with rapid compression of inference costs. Companies like Anthropic and OpenAI are betting on trust, safety, and enterprise contracts to maintain premium pricing, but the timeline is uncertain.
Chinese AI models dominated the top 6 usage on OpenRouter for last week and this week, with the free version of Tencent Hunyuan3 ranking first, Xiaomi MiMo-V2.5 and DeepSeek-V4-Flash in the lead, reflecting the significant increase in influence of domestic models in the open-source community.
A researcher demonstrates how to backdoor an open-weight coding model in under $100, raising concerns about trust in Chinese open-weight models like GLM 5.2.
A price tracker found that GLM-5.2 and Tencent's Hy3 quietly changed prices multiple times in a week, highlighting volatility in Chinese LLM pricing.
A test of four new frontier AI models (MiMo-V2.5-Pro, MiniMax M3, Mercury 2, LongCat-2.0) using riddles that require genuine reasoning rather than pattern-matching reveals that while most models perform reasonably, LongCat-2.0 repeatedly generates fabricated information with false confidence.
A user describes their fully local AI stack using multiple hardware devices running Chinese models like GLM, Qwen, and Kimi, claiming 87% cost savings compared to frontier models like GPT-5.5 and Opus 4.8, while noting plans to self-host video generation.
A list of Western companies including Lindy, Cursor, Coinbase, Shopify, Airbnb, Uber Eats, Siemens, and ChapsVision that are moving AI workloads to Chinese models such as DeepSeek V4, Kimi K2.5, GLM-5.2, and Qwen.
Xiaoyin Qu argues that American and European enterprises will ditch OpenAI and Anthropic in favor of Chinese models because they can be hosted on their own GPUs and fine-tuned with proprietary data, offering more control and compliance.
Discusses the potential for Chinese open source AI models to become the dominant option in the near future.
The author argues that Anthropic's call for a pause in AI development is a cover story to eventually ban Chinese AI models under the guise of security, as Chinese models like Deepseek, Minimax, Kimi, and Qwen are catching up at a fraction of the cost.
Analysis of OpenRouter data shows that Chinese AI models have become the most used in Kilo Code's coding agent, accounting for 58% of token usage, challenging the dominance of Claude and GPT due to lower cost and longer context windows.
Qwen 3.6 Max Preview launched on Qwen Chat website, achieving the highest AA-Intelligence Index score (52) among Chinese models, with uncertainty about whether it will be open source.