China AI

AIHuawei unveiled the Atlas 960E superpod, which links up to 4,096 NPUs using near-packaged optics (NPO) and claims eight exaflops at FP8 precision and up to one petabyte of high-bandwidth memory.

AIInferact, with Red Hat AI, NVIDIA, and Huawei, co-led an optimization effort that makes Kimi K3 on vLLM 2.2–2.8× faster.

AIMiniMax has become an AI partner to Singtel's newly launched AI Pass, supporting Singapore's SkillsFuture AI Subscription initiative.

AIOpenBMB's MiniCPM5-2B is a 2.5B-parameter model with native 128K context, offering hybrid Think and No-Think modes in one checkpoint.

AIVC-Attention, developed by Nunchux AI and collaborators, accelerates attention for MiniMax-H3 without retraining.

AISenseTime released its SenseNova U1.5 technical report, describing an open-source 8B native MoT unified model that connects understanding and generation through shared attention.

AIPruna AI has introduced P-Video-2-Pro, built on MiniMax H3, which generates 5-second videos at 480p or 720p in seconds. The model is 50% off for one week, starting at $0.01 per second.

AIBaidu's Apollo Go plans to apply for commercial operation of its autonomous driving service in Hong Kong, citing the HKSAR Government's support in its first Five-Year Plan and the 2026 Policy Address.

AIBaidu's Apollo Go robotaxi is conducting fully driverless testing on Hong Kong's roads. The post invites viewers to see the vehicles in action but gives no further details on scale, timing, or partners.

AIinclusionAI has released Ming-Image-0.1-Design-Layer on Hugging Face, a model that decomposes a flattened design image into a requested number of RGBA layers using an image and a layer plan.

AIHuawei unveiled the Ascend 960 SuperPoD at HUAWEI CONNECT 2026, supporting up to 4,096 NPU cards, 8E FP8 computing power, and 1PB of HBM capacity.

AISenseTime chief scientist Lin Dahua argues that native multimodal AI, which processes language, vision and other information in one shared model, is essential for AI to move beyond coding into industries and the physical world.

AIinclusionAI has released Ming-Image-0.1-Design, a 6B text-to-image model for UI, infographics, and posters that outputs RGBA images with transparent backgrounds.

AIZ.ai reports that GLM-5.3 helped build and optimize the inference infrastructure for GLM-5.3-Flash.

AIAlibaba's Wan3.0 video model now produces a single 30-second shot directly, up from 15 seconds and a year ago's 5-second clips.

AIOpenBMB says Dubedo is the kind of product it hoped VoxCPM2 would enable, citing speaker-aware cloning, multilingual generation, and an editing studio.

AIFountain 0's new feature film ODYSSEUS: The Fall, directed by Ash Koosha, is now available in full with every shot generated by Kling 3.0.

AIQwen's Wan 3.0 video model produced a complete fantasy fight between a mage and a swordswoman, with strong timing, poses, and complex layouts that the post calls anime-ready.

AIStepFun has published a Hugging Face Space demo, API documentation, and a thesis for StepAudio-3-Music. The post contains only links, so no features, benchmarks, or pricing are stated in the source.

AIDeepSeek engineer Shengyu Liu, who wrote DeepSeek-V4.1-Flash's attention kernels, published an essay arguing that AI industrializing software production may leave students without the skills to understand and plan whole projects.

AIPeking University's Yu Zhaofei team and the Beijing Academy of Artificial Intelligence (BAAI) released Brainμ-Spike, a small convolutional network for spike camera image reconstruction that is paired with the Brainμ model.

AIinclusionAI has published Realtime-Venus on Hugging Face with two 9B checkpoints: Realtime-Venus-Omni for audio-visual interaction and Realtime-Venus-Audio for audio-only conversation.

AITencent's WeVisDoc-4B, fine-tuned from Qwen3-VL-4B-Instruct, converts page images into structured Markdown with LaTeX formulas and HTML tables.

AITencent's WeVisDoc-4B, fine-tuned from Qwen3-VL-4B-Instruct, scores 95.38 Overall on OmniDocBench v1.6 and 75.54 mean Overall across three PureDocBench tracks.

AIHiDream.ai launched vivago R1, a content creation agent, globally, with a domestic version upgrade.

AIv0 now supports multiple AI models through Vercel's AI Gateway, spanning frontier, cheap, open, and fast options. Users can pick from models including Claude, GPT, Kimi, GLM, Grok, and DeepSeek to build their apps.

AIIn a Paint recreation test, GPT-5.6 Astra built the image from layered geometric shapes, while Qwen3.8 Max worked pixel by pixel.

AIMiniMax Design lets a single brief drive a full production workflow on one canvas, with the Astra agent working in Blender through an official connector to build the scene and camera direction.

AINovita AI has open-sourced Chord, a W4A16 MoE CUDA operator with BF16 activations, INT4 weights and group-32 scales, built for Kimi K2.x serving shapes.

AIThe vLLM team trained a DSpark speculative decoding draft model for Kimi K3, a 2.8T-parameter model, using the Speculators library on GB300 NVL72 hardware.

Why it matters: The post shows how hidden-state extraction and Mooncake transfers let a 2.8T-parameter model's speculator be trained across multiple nodes, a reusable pattern for similar setups.

AIMiniMax H3 with SGLang-Diffusion and VDN-H3 generates 14.4 seconds of 768p video in 9.0 seconds end-to-end after warmup on 8× B200 GPUs.

AIDario Amodei's essay "We Must Pace the Frontier" proposes slowing capability gains, starting with independent evaluators inside AI companies and extending to international coordination including China.

AIRadixArk's SGLang-Diffusion, paired with VDN-H3, generates 14.4 seconds of 768p video in 9.0 seconds on 8× B200 GPUs.

AISenseNova Skills is an open-source agentic office skill suite from SenseTime that combines Deep Research, Data Analytics, PPT Generation, and Infographic Generation.

AIIntern Large Models introduces Intern-S2-397B, a multimodal foundation model built for long-horizon scientific research and scientific agents.

Why it matters: The post pairs a new open multimodal model with benchmark tables against named Qwen, DeepSeek, Kimi, GLM, GPT, Gemini, and Claude models, letting readers compare scientific and agentic results directly.

AITencent released Simple-Attention-Sparsification (SAS) gate checkpoints for Qwen3-4B, Qwen3-8B, and Qwen3-14B, which learn to rank and select KV blocks using continuous gates optimized with the language-modeling loss.

AIBaidu's Miaoda no-code platform has upgraded with enhanced AI agents for design, app generation, and testing.

AIMiniMax highlighted open-source community progress on its H3 video generation model, which it built with native stereo audio and multimodal reference control.

AIHailuo AI promotes bringing AI tools into professional 3D production workflows.

AIAt Guotai Junan Securities' 2026 Autumn Conference, SenseTime's Head of Capital Markets Philip Wong laid out three shifts reshaping AI: from single-modal to native multimodal, from token consumption to task delivery, and from single-point models to system-level full-stack capabilities.