At Apsara 2026, Alibaba debuted Qwen3.8-LiveTranslate with LAAL latency cut from about 2.8 to 2.3 seconds and unveiled the Qwen-Audio-3.1 family spanning ASR, TTS, Realtime and TTS-Next cinematic ...
At Connect 2026, Huawei Cloud said AgentArts already serves 100+ enterprises alongside openJiuwen, Agentic MaaS, and Industry AI Foundry assets for production agents. At HUAWEI CONNECT 2026 in ...
Want to read in a language you're more familiar with? Tencent WeChat open-sources WeKnora, a RAG-to-agent knowledge platform with 25k+ GitHub stars, plus an official DeepSeek Harness plugin that adds ...
DeepSeek released DeepSeek Harness, an MIT-licensed agent runtime in developer preview where models, tools, sessions, sandboxes and UI are Cordis plugins, with npm packages advancing through the ...
Huawei lifted Ascend 950DT indicative quotes past about 250,000 yuan and Cambricon raised planned 690 pricing 20%–30%, Reuters and Bloomberg report, as HBM ...
Security researcher Chaofan Shou said he obtained about 6TB of invocation logs from a Chinese large-model relay and found SSH keys, cloud credentials and tokens tied to enterprises. The report is ...
Want to read in a language you're more familiar with? Huawei's Kirin 9050 Pro commercializes LogicFolding under Tao's Law, raising transistor density about 55% and cutting power in key workloads. The ...
Qwen has released Qwen-Drive-1.0-4B, an Apache 2.0 autonomous-driving vision-language model on a Qwen3.5-4B base with 3D perception and trajectory planning. SFT and RL Planning Expert variants are ...
XPeng has switched on an automated IRON humanoid production line with more than 80% core automation, as a finished robot walked off under its own power. Mass production is targeted by the end of 2026, ...
PIF-backed HUMAIN launched an Arabic frontier model on MiniMax's open MoE stack—showing how sovereign AI projects are turning Chinese open weights into national starting points. HUMAIN-M3 keeps M3's ...
He Tingbo's latest paper cites Kirin 2026 silicon: ~55% higher transistor density with sharp NPU/GPU/CPU power cuts at matched performance—arguing LogicFolding can cool, not melt, stacked logic. The ...
Want to read in a language you're more familiar with? China ships most of the world's humanoids, but analysts say the milestone is task generality—not demos—until robots do ~80% of novel jobs from ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results