2+ day, 23+ hour ago   (1808+ words) Just two days after DeepSeek tested the intermediate version of V4.1 Flash, V4.1 Flash was officially launched. This MoE model with a total of 552B parameters adopts a brand new asymmetric Causal-Encoder-Decoder structure. It only has 8B input activation and 16B output activation, with inference…...

3+ day, 19+ min ago   (982+ words) ZheDongXi reports on September 10 that DeepSeek has just officially released DeepSeek V4.1 Flash. Built with a brand-new architecture, this 552B-parameter MoE model is the smallest member of DeepSeek's new model architecture family, outperforming a host of flagship models including DeepSeek V4 Pro…...

3+ day, 1+ hour ago   (580+ words) DeepSeek probably believes this is a very simple and clear deal. According to internal and external tests, V4.1 Flash has fully surpassed V4 Pro in performance, cost, speed and total latency. The internal beta notice also mentioned that the new model adopts…...

3+ day, 8+ hour ago   (1647+ words) Douyin has taken the throne of the mobile internet market. Nomura's latest research report, citing data from QuestMobile, states that in July 2026, Douyin's total usage duration surpassed WeChat for the first time. The duration share of the ByteDance ecosystem among…...

3+ day, 2+ hour ago   (296+ words) • Original "one brain for multiple forms": the same "brain" is naturally adapted to multiple ontology configurations such as muscle joints, flexible single arms, flexible dual arms, flexible dexterous hands/androids and their corresponding task requirements; • 1B+ massive dynamic parameter combinations: relying…...

3+ day, 8+ hour ago   (1733+ words) If you only look at the recent moves of AI companies, no one would feel that this frantic competition is showing any sign of slowing down. New models are still being released one after another, company valuations are getting higher…...

4+ day, 5+ hour ago   (1682+ words) Careful viewers may notice that whether it is racing, obstacle course, high jump, weightlifting or long jump, there are always one or two operators standing behind the robots, holding remote controllers in their hands to issue instructions to the robots....

4+ day, 9+ hour ago   (1052+ words) DeepSeek officially stated that, "DeepSeek V4.1 Flash adopts a new model architecture, features native multimodal support, stronger capabilities, faster speed, and lower cost." There is no technical report or performance parameter sheet, but the information released by the official is already…...