News
ReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior ReleaseReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior Release
EveryLLM

Lab city, China

Shenzhen

Huawei's PanGu and Ascend chips and Tencent's Hunyuan: China's hardware and platform giants.

Shenzhen, across the border from Hong Kong, is home to Huawei and Tencent. Huawei builds the PanGu models and the Ascend AI chips that many Chinese labs turned to under US export controls; Tencent builds the Hunyuan models and the Yuanbao assistant, reaching users through WeChat. The city's electronics industry also makes it a centre for on-device AI.

Huawei

Huawei launched its PanGu models in 2021 and described PanGu-Σ, a sparse model with over a trillion parameters, in 2023. PanGu models are aimed mainly at industry, for tasks such as weather forecasting, mining and finance, and run on Huawei's Ascend chips and its MindSpore software. In 2025 Huawei released some PanGu weights openly as openPangu.

Tencent

Tencent announced Hunyuan in September 2023 and released Hunyuan-Large in November 2024, an open mixture-of-experts model with 389 billion parameters, 52 billion of them active. Its Yuanbao assistant added DeepSeek-R1 alongside Hunyuan in early 2025, a sign of how Chinese platforms mix in-house and outside models. Tencent also builds 3D and video generation models under the Hunyuan name.

Models of Shenzhen

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.