China AI news, from the source
Top stories
Updated Oct 16, 2026 Share this briefingToday's brief is being compiled. Verified Events keep landing in the live update ledger below and in the full Event History.
Browse the Event HistoryMore China AI news
Evidence-qualified product, pricing, policy, capital and deployment Events, ordered by when they happened. Research appears only after an impact review; unverified Signals stay separate.
20 of 52 stories · last 7 days
Newest first. Wider windows expand what is available; open Event History for the full period.
Today · Oct 16, 2026 15 stories
Tencent open-sourced DSpine, a parallel drafter for Qwen3-8B speculative decoding
Tencent released Qwen3-8B-DSpine-b16 on Hugging Face under Apache 2.0. The model is a parallel drafter for speculative decoding, achieving an average acceptance length of 7.83 at …
Why it matters Developers using speculative decoding can now test a higher-acceptance drafter for Qwen3-8B under Apache 2.0, potentially reducing serving cost per token.
Reportedly: Apple to launch Apple Intelligence in mainland China via iOS 27.3 in December with Baidu and Alibaba as suppliers
Apple plans to launch Apple Intelligence in mainland China via iOS 27.3 in December, IT Home reported, citing Bloomberg. The rollout covers iPhone, iPad, and Mac with AI features …
Why it matters Apple's entry into mainland China's AI market with local partners Baidu and Alibaba will require those partners to meet Chinese regulatory requirements, potentially shifting competitive dynamics for…
Reportedly: Tencent Cloud open-sources TeamAI, a Git-based tool for sharing skills across agents
Tencent Cloud open-sourced TeamAI, an internal Git-based team AI collaboration tool that stores skills, work rules, tool configs, and project knowledge in a Git repo for review, u…
Why it matters Development teams using multiple AI coding agents gain a centralized way to manage and version shared skills and rules, reducing per-developer setup and inconsistency.
Reportedly: Tencent is developing a standalone 'Yuanbao Assistant' app for personal agent tasks
Tencent is developing a Personal Agent product called 'Yuanbao Assistant' with a standalone app, IT Home reported, citing people familiar with the matter. The app would run long-c…
Why it matters Tencent's reported move into a standalone personal agent app would give it a direct competitor to ByteDance's 'Xiaodou', potentially shifting consumer AI assistant competition toward long-term…
Reportedly: Tencent Yuanqi to shut down on November 9, 2026, making created agents unavailable
Tencent Yuanqi will stop service on November 9, 2026, making all created agents unavailable, IT Home reported. From October 29, 2026, new agent creation and API key distribution w…
Why it matters Developers and businesses using Tencent Yuanqi for agent deployment must migrate to alternative platforms before November 9, 2026, or risk losing access to their agents and data.
Reportedly: China proposes adding AI, biotech, and solar technologies to export control list
China's Ministry of Commerce and Ministry of Science and Technology issued Announcement No. 48 of 2026, proposing to add 40 key technologies in AI, biomedicine, and photovoltaics …
Why it matters Chinese AI, biotech, and solar companies face new compliance duties if the proposal is adopted, as exporting the listed technologies would require licenses or be prohibited, potentially affecting…
Reportedly: 4Paradigm and RiseUnion Open-Source ModelSphere for Heterogeneous LLM Serving
4Paradigm and RiseUnion open-sourced ModelSphere, an LLM deployment tool unifying cluster serving, KV caching, and AutoTune under Apache 2.0. It claims support for Ascend, Iluvata…
Why it matters Chinese enterprises using domestic AI chips gain a unified open-source deployment layer, potentially lowering the cost and complexity of serving LLMs on non-NVIDIA hardware.
Reportedly: AgiBot starts new funding round at about RMB 50 billion valuation
AgiBot is raising RMB 1–1.5 billion in a new round nearing completion at a valuation of about RMB 50 billion, IT Home reported, citing Caijing. The round means AgiBot cannot compl…
Why it matters AgiBot's new funding round delays its IPO until at least 120 days after the financing closes, which affects investors' exit timeline and may push the listing into 2027.
Reportedly: DeepSeek web and API services suffer partial outage with 'server temporarily unavailable' errors
DeepSeek's web and API services experienced a partial outage on October 15, 2026, with users seeing 'server temporarily unavailable' errors. The official status page confirmed the…
Why it matters DeepSeek users and API customers outside China face service disruption, potentially impacting applications and workflows dependent on the platform.
inclusionAI released LemonDecider, a 4B decision model with explicit abstention
inclusionAI released LemonDecider, a 4B-parameter decision model derived from Qwen3.5-4B, on Hugging Face under Apache-2.0. It uses a 255-way single-token readout for choice, noul…
Why it matters Developers outside China can now use a specialized decision model from Ant Group's inclusionAI, reducing the need to adapt general-purpose LLMs for structured choice tasks.
Reportedly: OpenCode makes DeepSeek V4.1 Flash free with 200K context
OpenCode announced DeepSeek V4.1 Flash is now free, with the free tier supporting a 200K context window, down from the 1M version, IT Home reported. The move follows earlier quota…
Why it matters Developers using OpenCode can now access DeepSeek V4.1 Flash without quota limits, lowering their cost for tasks within 200K context.
Reportedly: StepFun's Step 5 Preview Comes to Tencent WorkBuddy, Free to Use Exclusively Until October 25
StepFun's Step 5 Preview model is now available inside Tencent's WorkBuddy office agent, with an exclusive free-use window running from October 12 to October 25, Pandaily reported.
Why it matters Tencent WorkBuddy users gain free access to StepFun's Step 5 Preview model until October 25, potentially reducing their AI costs during the promotional period.
Reportedly: Xiaomi open-sources MiDashengLM-Spatial, an end-to-end spatial audio understanding large model
Xiaomi released MiDashengLM-Spatial, an open-source end-to-end spatial audio understanding large model, with weights and inference code under Apache 2.0. It builds on MiDashengLM-…
Why it matters Developers outside China gain access to an open-source spatial audio model that can be integrated into smart home and automotive applications, potentially reducing the cost of building spatial audio…
Reportedly: Kuaishou Kling 4.0 video model fully launched, supports up to 30-second generations
Kuaishou's Kling AI announced the full launch of Kling 4.0 on klingai.com, with single video generations up to 30 seconds. 4K and 1080p 10-bit HDR output is coming soon, and API a…
Why it matters Kuaishou's full launch of Kling 4.0 puts pressure on competing video generation providers to match its 30-second duration and upcoming HDR output, potentially shifting enterprise expectations for…
Reportedly: Meta releases Muse Spark Open, an 816B-parameter open-weights multimodal model
Meta released Muse Spark Open, an open-weights model with 816B total parameters, 42B active per token, and 1M-token context. It supports multimodal input and targets coding and ag…
Why it matters Meta's release puts a US open-weights model directly against Chinese open-weights models on public benchmarks, where it scores below several Chinese models on Artificial Analysis' intelligence index.
Yesterday · Oct 15, 2026 5 stories
Reportedly: DeepSeek open-sources operator toolchain with native Huawei Ascend support
DeepSeek released open-source versions of its operator development tool TileLang and five core compute/communication libraries (DeepGEMM, DeepEP, FlashMLA, TileKernels, DeepSelect…
Why it matters Developers using Nvidia's CUDA ecosystem now have an open-source alternative that runs on Huawei Ascend NPUs, potentially lowering the cost and complexity of porting AI workloads to Chinese-made…
Tencent Hunyuan open-sources OmniCapBench audio-visual captioning evaluation framework
Tencent Hunyuan released OmniCapBench, an open-source evaluation framework for fine-grained audio-visual captioning, on GitHub. The repository includes an annotation pipeline and …
Why it matters Researchers and developers evaluating audio-visual captioning models gain a reproducible, open-source benchmark that localizes errors to specific units, reducing reliance on opaque global LLM judges…
Reportedly: ByteDance's Doubao Phone Assistant blocked from GUI automation by 151 apps including WeChat, Taobao, and Meituan
ByteDance's Doubao Phone Assistant will not perform GUI automation on 151 apps that declined under the Screen Automation Operation Statement Protocol, including WeChat, Taobao, an…
Why it matters App developers gain a formal opt-out from Doubao Phone Assistant's GUI automation, preserving control over their user experience and security.
Reportedly: Manus AI releases Manus 2.0 with Cascade framework, cloud computers, and Cue personal assistant
Manus AI released Manus 2.0, introducing the Cascade agent framework that cuts task token consumption by 23.2%, completion time by 28.2%, and running costs by 32%, GeekPark report…
Why it matters Manus AI's 2.0 release directly pressures competing agent platforms by bundling cloud computing, automation, and a personal assistant into one ecosystem, potentially reducing the need for users to…
Reportedly: Shanghai AI Lab Open-Sources Intern-Decision, Small Models That Output Decisions, Not Text
Shanghai AI Lab open-sourced Intern-Decision, a family of small models in 0.8B, 2B and 4B sizes that output decisions, scores and yes/no answers with probabilities. MetaX shipped …
Why it matters Developers outside China gain access to open-weight decision-focused models that can be fine-tuned or deployed locally, potentially reducing reliance on large general-purpose LLMs for classification…
Reportedly: DiffuSpace raises 500M yuan, Huawei and Horizon Robotics among investors
Shenzhen-based DiffuSpace is raising a 500 million yuan (approx. $70M) funding round, with Huawei and Horizon Robotics among the investors, TMTPost reported, citing people familia…
Why it matters Huawei and Horizon Robotics would gain exposure to an alternative language model architecture, potentially reducing dependence on autoregressive models for edge and automotive applications.
Reportedly: Kunlun Tech releases Matrix-Game 4.0, the first video large model game running on a single consumer GPU, and open-sources it
Kunlun Tech released Matrix-Game 4.0, the first video large model game running on a single consumer GPU, achieving 480p at 60 FPS with under 150 ms latency on an NVIDIA RTX 5090. …
Why it matters Game developers outside China can now access an open-source system that runs a video large model game on a single consumer GPU, potentially reducing the hardware barrier for prototyping AI-driven…
AgiBot open-sources VINE, a value-gradient fine-tuning method for flow-matching policies
AgiBot released the VINE codebase on GitHub under Apache License 2.0. VINE enables stable end-to-end value-gradient fine-tuning of flow-matching policies by reconstructing interpo…
Why it matters Robotics and embodied AI developers outside China gain access to a new open-source method for fine-tuning flow-matching policies with value gradients, potentially reducing the engineering effort…
Reportedly: ByteDance's Doubao phone assistant faces rejection from 155 apps over screen automation
ByteDance's Doubao phone assistant saw 155 third-party apps reject its AI screen automation after the SAEP protocol's 30-day public comment period ended on October 14, GeekPark re…
Why it matters App developers gain a formal mechanism to block Doubao's screen automation, protecting their traffic and security; ByteDance must now negotiate access app by app, which could slow Doubao assistant's…
No additional verified events in this window. Try a wider period or open Event History.
View All eventsNew Chinese AI models
All models- XiaomiMiMo-V2.6-ProSep 21, 2026 · Open weights
- DeepSeekDeepSeek-V4.1-FlashSep 10, 2026 · Open weights · Global API
- Zhipu AIGLM-5.3-FlashAug 26, 2026 · Open weights · Global API
- AlibabaQwen3.8-Flash-NextAug 26, 2026 · Open weights · Global API
- Zhipu AIGLM-5.3Aug 18, 2026 · Open weights · Global API
- AlibabaQwen3.8-27BAug 14, 2026 · Open weights · Global API
Explore China AI
- China AI modelsAccess, open weights, and licenses across Chinese labs.
- Chinese AI companies73 frontier labs, big tech, robotics and AI chip makers.
- China AI policyRules, algorithm filings, labeling standards and export controls.
- Attention IndexEnglish and Chinese source attention.
- China AI ExplainersEssential primers on terminology, frontier startups, and compliance.
- China AI LinesFive strategic themes: open weights, price wars, robotics, chips and policy.
Your reading list
Save stories. Free, across devices.