Foreign media report that Tesla's Optimus V3 is running into trouble with its robotic hands, its production line and its training data, and that the company now builds a few hundred units a week. XPeng's robot has cleared supply chain factory audits, while Changan has set its mass production target for 2028.
The model card lists 890 bytes of cache per token, about a quarter of what V4-Flash uses, and HBM demand at a quarter of the previous generation.
Out on 10 September 2026: a multimodal Mixture-of-Experts with 552 billion parameters fires 8 billion per token on the way in and 16 billion on the way out. The architecture changed, not just the scale.
China's Cyberspace Administration says 1,112 generative AI services had completed filing as of August 31, 2026. The same week, the AI Safety Governance Framework 3.0 came out during Cybersecurity Week.
In the model card's table, V4.1-Flash scores 90.6 on Terminal-Bench 2.1 and 74.2 on DeepSWE v1.1, matching or beating Opus-5.0, GPT-5.6 Sol, K3 and GLM-5.3.
Reuters and The Information report that DeepSeek has hired CITIC Securities to prepare a STAR Market listing and is pushing a second funding round targeting 50 billion yuan. Annualized revenue is said to have passed 1 billion US dollars.
MiMo-V2.6-Pro has 1.02 trillion parameters, 42 billion of them active. In an open live training run it scored 46 points on the Artificial Analysis index and took first place among open models.
At AICC 2026, an IDC report split the compute gap into two layers: the ceiling on capability and the scale of intelligence. By 2030 the shortfall could reach 380.9 billion dollars, and stacking cards alone can no longer cover the different loads of Prefill and Decode.
Xiaomi streamed an entire reinforcement learning run live, and MiMo-V2.6-Pro took the top open-source spot with 1.02 trillion parameters and 42 billion active. DeepSeek also released a vision experiment under the MIT license.
A new DeepSeek paper signed by Liang Wenfeng describes DSec, an agent training system that produces more than 5,000 sandboxes a second, reaches 3 million in a single day and peaks at 380,000 running at once.
On the Artificial Analysis Intelligence Index, Xiaomi's MiMo-V2.6-Pro and the closed-source Grok 4.7 both score 46. But one task costs about 29 times more on Grok 4.7. Cost is turning into a benchmark dimension of its own.
Liu Sheng leads the main Attention operator on DeepSeek V4.1. He wrote "I Had to Bury My Talent Yesterday"; the post hit No. 1 on Zhihu's trending list within a day. Drawing on his public writing, we look at how AI coding is changing the job.
Kunlun Tiangong showed off Mureka V8 at a concert in a music club and says the model beat Suno V5 in tests with musicians. The company's CEO talks about a new genre, not just another tool.
Long context usually runs into memory, not weights. DeepSeek V4.1 Flash squeezes its KV cache to about 890 bytes per token, cutting HBM demand to a quarter of the previous generation and SSD demand to an eighth.
V4 Pro was due to go offline on September 14. On September 11, DeepSeek said it would keep serving API calls, with billing unchanged. The earlier plan had routed all V4 Pro requests to V4.1 Flash.
DeepSeek released V4.1 Flash on September 10 Beijing time. The new MoE model carries 552 billion parameters but activates only 8 billion on input, and it cuts KV Cache demand on HBM to a quarter of what the previous generation needed.
WeChat Pay's AI card now works with DeepSeek Harness and OpenClaw. Once a user authorizes it, the card handles the whole flow from recommendation to payment inside a conversation. WeChat stresses main-account isolation and per-order confirmation.