AI AI Toolkit
🇨🇳 Updated Daily

China AI Dynamics

Latest progress and ecosystem development of domestic large models. Track Baidu, Alibaba, Tencent, ByteDance LLMs dynamics.

Domestic Focus
Enterprise
📅 100 articles
Updated 2026-11-22

月之暗面���Kimi)发布官方教程,详解 Kimi K3.1 Pro 的 3D 创作工作流程,说明参数化建模与网格化建模两种方式及对应工具选择。文章分享了 4 个案例的具体操作,包括在 Unreal Engine 中构建东方天宫、制作可 3D 打印的台灯、纯代码调用 Blender 生成建筑、用 Tripo 和 Blender 搭建微缩城市,并给出图胜于言、先搭灰模再细化等 5 个使用技巧

📰 公众号:月之暗面(Kimi) · 2026-10-30

Well-known AI tech blogger Kazik (@Khazix0918) shared on X sharp observations on China’s AI products and model developments, focusing on big-tech launches, open-source progress, and industry trends.

📰 X:卡兹克 (@Khazix0918) · 2026-10-29

Tencent Hunyuan (@TencentHunyuan) shared new progress in large models, multimodal capabilities, and AI applications, reflecting Tencent’s dual "enterprise plus consumer" AI push.

📰 X:腾讯混元 (@TencentHunyuan) · 2026-10-29

Ant Group's Ling-3.1-flash has about 560B total parameters with ~25B activated per token and a 1M context window, continuing the hybrid-linear architecture and raising the linear-attention ratio (7 KD

📰 公众号:蚂蚁百灵(Ling) · 2026-09-30

DeepSeek open-sourced infrastructure components for Huawei's Ascend compute platform, including the TileLang compiler, DeepGEMM, DeepEP, TileKernels, FlashMLA, and DeepSelect, mirroring its earlier NV

📰 公众号:DeepSeek(深度求索) · 2026-09-30

The author summarizes OpenAI DevDay 2026: the personal agent product Dots launched to ChatGPT Pro, Business Premium and Enterprise users with 4,000+ app integrations; new GPT-6.1 Sol went live at abou

📰 公众号:数字生命卡兹克 · 2026-09-29

Xiaomi released MiMo-V2.6 Pro and Flash, open-weight full-modal models trained via large-scale trial-and-error learning; Pro scores 46 on the Artificial Analysis Intelligence Index, the highest among

📰 X:Aravind Srinivas(Perplexity CEO) (@AravSrinivas) · 2026-09-23

Ant Group released the open-source Ming-Image-0.1-Design series with two 6B models: Design generates complete designs such as UI, infographics, and posters from text, while Layer decomposes a design i

📰 公众号:蚂蚁百灵(Ling) · 2026-09-23

An independent author shared a workflow that saves Codex quota: wrap your own server as a read-only, least-privilege, Feishu-OAuth MCP Server that GPT-6 Pro on the ChatGPT web app calls as a plugin, r

📰 公众号:数字生命卡兹克 · 2026-09-17

NSA FBI CISA alleged DeepSeek Moonshot AI Moonshot AI Alibaba MiniMax StepFun Z.ai 2024 model extraction said coding capability model model

📰 X:X.PIN (@thexpin) · 2026-09-09

NSA CISA FBI released AA26-251A alleged DeepSeek Moonshot AI Alibaba MiniMax Z. AI 2024 model model model

📰 X:Rohan Paul (@rohanpaul_ai) · 2026-09-09

Open AI shipped GPT Image 2.5 generation editing 15 image call interface released GPT-Image-2.5 Flare 50% GPT-Image-2.5 Sunburst

📰 公众号:卡尔的AI沃茨 · 2026-09-09

GPT-6 Astra inference Reasoning Effort model Ultra AI collaboration

📰 公众号:数字生命卡兹克 · 2026-09-09

GPT-6 Astra shipped users Blender Houdini Unity Aseprite Demo 3D Astra

📰 公众号:数字生命卡兹克 · 2026-09-07

MiniMax connected its H3 Max model at 768P and 480P to its open platform and MiniMax Design. Overseas developers have already used the APIs to build a Twitch livestream and a 24-hour 'AI TV station' t

📰 公众号:MiniMax(稀宇科技) · 2026-08-31

Doubao Work is the lowest-barrier path for enterprises to adopt agents but needs a Feishu login for full features; it can remotely control up to seven devices, run scheduled tasks, read local skills,

📰 公众号:卡尔的AI沃茨 · 2026-08-26

GLM-5.3 API is live today, excelling at complex coding, defensive cybersecurity, and long-horizon tasks. It scores 60 on the Artificial Analysis Intelligence Index, on par with undisclosed internal fl

📰 公众号:智谱(GLM) · 2026-08-19

Qwen has open-sourced the Qwen3.8 model series. Qwen3.8-27B is a native multimodal (image/voice/text) dense model whose 27B parameters already reach the level of Qwen3.7-Plus, with native 262K context

📰 X:通义千问 / Qwen (@Alibaba_Qwen) · 2026-08-14

A beginner-friendly walkthrough of the DeepSeek Harness, an agent framework that orchestrates models, tools, skills, and loops into reusable workflows. The piece covers what the Harness is, its core c

📰 公众号:数字生命卡兹克 · 2026-08-13

Xiaohongshu (RED) dots team open-sourced dots.tts, a 2-billion-parameter fully continuous end-to-end autoregressive speech synthesis model, achieving the best average content accuracy and average spea

📰 公众号:小红书技术(dots.llm) · 2026-08-13

Ant Ling (Bailing) has open-sourced Ling-3.0-tiny, a native hybrid reasoning model with 7.9B total parameters that activates only 1.3B parameters during inference, and simultaneously releases three ve

📰 公众号:蚂蚁百灵(Ling) · 2026-08-11

Based on Xiaowei, WeChat launched internal tests of Moments AI writing-assist and AI commenting: the former generates three Moments captions from an image and written text, while the latter lets you l

📰 公众号:数字生命卡兹克 · 2026-08-11

The author spent 54 hours building and freely releasing LatentRank, a comprehensive AI model leaderboard that aggregates multiple trusted rankings. It uses the Bradley-Terry pairwise comparison algori

📰 公众号:数字生命卡兹克 · 2026-08-10

ModelBest's OpenBMB unveiled AMNESIAC at the #BuildSmall hackathon, a reverse Turing test interactive game where players must convince an AI interrogator named A.M.N. that they are human. Powered by M

📰 X:面壁智能 OpenBMB (@OpenBMB) · 2026-08-06

The author compiled 10 open-source apps that improve the vibe coding experience, covering Mac notch customization (Atoll), window preview (DockDoor), a fast launcher (Raycast), thorough uninstallation

📰 公众号:卡尔的AI沃茨 · 2026-08-06