Deepseek releases V4.1-Flash, a multimodal model with 552 billion parameters that cuts KV cache memory to a quarter of its predecessor. On the DeepSWE coding benchmark, it narrowly beats Opus 5 and GPT-5.6 Sol, even though only 16 billion parameters are active per token. The model ships under the MIT license and targets much cheaper AI agents.

The article New Deepseek model V4.1-Flash cuts memory needs for AI agents appeared first on The Decoder.

Deepseek wants to put 160,000 Huawei Ascend-950DT chips into an Inner Mongolia data center for inference only, not training. It would be the largest known Huawei chip cluster. But production bottlenecks mean Huawei probably can't deliver for over a year.

The article Deepseek plans the largest known Huawei chip cluster with 160,000 processors in Inner Mongolia appeared first on The Decoder.

Deepseek has released V4-Flash-Vision-Exp, an experimental multimodal model that adds image understanding to V4-Flash's text capabilities. On the company's own multimodal agent benchmarks, it approaches Opus 4.8 and sometimes beats it.

The article Deepseek releases experimental Flash vision model that rivals Opus 4.8 on agent benchmarks appeared first on The Decoder.

Unitree Robotics’ initial public offering has reportedly been oversubscribed more than 5,000 times as investors scramble for shares in one of China’s best-known humanoid robot manufacturers. The Hangzhou-based company priced its IPO at 150.8 yuan ($22.4) per share, raising approximately $900 million and giving Unitree a valuation of around $9 billion, according to CNBC. Demand […]

Chinese robotics company Unitree has priced its initial public offering at a valuation of approximately $9 billion, becoming the country’s first mainland-listed humanoid robot manufacturer and highlighting growing investor interest in AI-powered robotics despite rising geopolitical tensions. The Hangzhou-based company said it has priced its Shanghai STAR Market IPO at 150.8 yuan ($22.34) per share, […]