China AI

AIDeepSeek released open weights for V4.1-Flash, which Baseten now offers through its Model APIs.

Why it matters: The post explains how the CED architecture splits prefill and decode compute and cuts KV cache memory, which matters for coding agent costs.

AIDwarkesh Patel said a discussion with John, Beren, and Charlie questioned why Sonnet 5 and Opus 5 feel weaker than GLM 5.3, even though Anthropic could use raw logit distillation from Fable and train on Fable's environments.

AIBAAI introduced AREX, a research agent built on a 122B-parameter mixture-of-experts model with 10B active parameters.

AIMeituan's LongCat team hosted the Hugging Face team at its Shanghai office to discuss future model development and open-source plans.

AILM Studio now offers DeepSeek V4.1 Flash, hosted in the United States with Zero Data Retention enabled by default. The model is accessible through a dedicated LM Studio model page.

AILM Studio has launched its Bionic version, now live for users.

AIMiniMax's Hailuo AI account announced an upgrade to its design platform, directing users to design.minimax.io. The post provides no details about the features or changes included in the upgrade.

AIMiniMax Design has released a major upgrade that adds GPT-6 Astra, new MCP integrations for Blender, Photoshop, and After Effects, and collaborative projects for teamwork.

AINathan Lambert has compiled a reading list of open-model writing covering why labs release open weights, the open-versus-closed debate, and US-China competition, last updated 15 September 2026.

AIShanghai Artificial Intelligence Laboratory has released Atria Dawn Preview, an agentic model built on the 744B-parameter MoE GLM-5.2 foundation model, with a 256K context window.

Why it matters: The release page gives a full benchmark table against named rivals and setup steps for Codex and Claude Code, useful for anyone evaluating agentic models.

AIOllama says DeepSeek-V4.1-Flash is now fully rolled out on its cloud, hosted in the US and Europe.

Why it matters: The post names the hosting regions, zero data retention policy, and pricing parity with the DeepSeek API, which matter for teams weighing cloud access to this model.

AITencent Hunyuan posted that much more AI is coming, with no specific model, date, or feature named.

AIDeepSeek-V4.1-Flash is now rolling out to Pro plan subscribers, according to Ollama. The background post says it is first available on Ollama's cloud for Max and Team accounts, with capacity being added to reach all subscribers.

AIChina's Ministry of Industry and Information Technology (MIIT) has issued the Implementation Plan for the "AI+Software" Special Action, notice No.

AIThe Cyberspace Administration of China, the Ministry of Agriculture and Rural Affairs, and the Ministry of Industry and Information Technology issued the Digital Village High-Quality Development Action Plan (2026–2030).

AIThe Cyberspace Administration of China, the Ministry of Agriculture and Rural Affairs, and the Ministry of Industry and Information Technology jointly issued the Digital Rural High-Quality Development Action Plan (2026–2030).

AIChina's Cyberspace Administration of China has issued the Digital Rural High-Quality Development Action Plan (2026–2030), setting a target that by 2030 digital rural development achieves notable results.

AIOllama announces that the DeepSeek V4.1 Flash model is available on its library, with a link to the model page. The post provides no further details about capabilities, parameters, or pricing.

AIOllama is rolling out DeepSeek-V4.1-Flash on its cloud, starting with Max and Team accounts. The company says it is quickly adding capacity to extend access to all subscribers.

AISebastian Raschka says DeepSeek V4.1 contains a major architecture overhaul using an encoder-decoder setup, and he argues it could have been named V5.

AILewis Tunstall joked that DeepSeek released a model just after Hugging Face's TRL library removed support for seq2seq models. The post is a lighthearted remark with no further technical details or figures.

AITencent Hunyuan has released AuK, an open-source foundation model for unified speech generation and editing that takes natural-language instructions and reference audio.

Why it matters: The release combines speech generation and editing under one natural-language interface, and its 4-step AuK-Flash variant reports about 4.5 times faster inference under matched conditions.

AIRadixArk links to a blog post on DeepSeek V4.1 at lmsys.org. The post text provided contains only the URL, so no further details on the model or its performance can be reported.

AIRadixArk has published a cookbook on its Miles documentation site covering how to run DeepSeek V4.1 Flash.

AIRadixArk says Miles brings day-0 RL support to DeepSeek-V4.1-Flash, with SGLang providing inference support.

AISGLang and Miles ship day-0 inference and RL support for DeepSeek V4.1 Flash, with weights now available.

AIDeepSeek says it will work with the open-source community on inference support for DeepSeek-V4.1-Flash and explore more deployment options.

AIDeepSeek says its more efficient V4.1-Flash architecture lets it serve more users at lower cost and pass the savings on through lower API prices.

AIDeepSeek says V4.1-Flash is now live on its API with native multimodal support, accessed through the model name deepseek-flash.

AIDeepSeek says its V4.1-Flash model needs only 1/4 the HBM and 1/8 the SSD storage for its KV cache compared with the previous generation.

AIDeepSeek has introduced a 552B-parameter MoE model built on a new Causal Encoder–Decoder architecture, activating 8B parameters for input and 16B for output.

AIDeepSeek introduces DeepSeek-V4.1-Flash, the smallest model in its new architecture family, with native visual understanding.

AIDeepSeek officially released DeepSeek-V4.1-Flash, the smallest model in its new architecture family, with native multimodal visual understanding.

Why it matters: The release lists benchmark results alongside API model-name changes and retirements, so developers can check both capability claims and migration steps.

AIBAAI released open-source resources for three AIDD projects on Hugging Face: EPT, an equivariant pretrained transformer for unified 3D molecular representation learning, and UniPath, a learnable-time flow matching method for crystal structure and energy prediction.

AIDeepSeek released DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts model with 552B backbone parameters and support for contexts up to one million tokens.

Why it matters: The report shows KV cache per token falling to about one quarter of DeepSeek-V4-Flash, a concrete tradeoff between long-context serving cost and benchmark results.

AIGenspark and Fireworks Lab post-trained the open-weight MiniMax M3 into Gen-1 Slides, a model that plans, writes, and checks slide decks end-to-end.

Why it matters: The post explains a post-training pipeline with reward design, curriculum, and numerical fixes, showing how a cheaper model was tuned toward a frontier quality bar.

AIKling AI's official account invites followers to stop by its booth to see the product in action. The post offers no details about the event, location, or demonstrations.

AIBaidu's Apollo Go says its all-electric RT6 robotaxi is now operating on roads across multiple countries around the world. The post is a brief announcement with no specific figures on locations, fleet size, or launch dates.

AIKimi has launched Remote Control in Kimi Work, letting users leave the app running on their computer and keep tasks moving from their phone. The post frames this as enabling users to work anywhere, anytime.

AIAt Arm Everywhere China on September 8, 2026, Arm launched products spanning data center CPUs, mobile compute subsystems, and robotics platforms.