News
EveryLLM

Chinese

Built for Chinese as well as English: Qwen, DeepSeek, GLM, Ernie, Kimi. 39 models on EveryLLM so far.

Chinese-language models are built for Chinese as well as English, and the strongest now compete with the best models anywhere. China's labs, including Alibaba, DeepSeek, Zhipu (Z.ai), Moonshot AI, Baidu, ByteDance, Tencent, Huawei, MiniMax and iFlytek, released so many models in 2023 that Chinese media called it a battle of a hundred models, and families such as Qwen and DeepSeek are among the most downloaded open models in the world.

Rapid growth

Baidu's Ernie Bot opened for testing in March 2023 and to the public that August, soon after China's rules for generative AI services took effect on 15 August 2023; services must be registered and their answers must follow state content rules. Tsinghua spin-off Zhipu had released the open GLM-130B in 2022. DeepSeek's V3 (December 2024) and R1 (January 2025) drew worldwide attention for strong results at a reported low training cost.

Chips, openness and censorship

US export controls since October 2022 limit China's access to the most advanced Nvidia chips, pushing labs towards efficiency, such as mixture-of-experts designs and lower-precision training, and towards Huawei's Ascend chips. Supporters of the controls see them as a security measure; critics argue they spur Chinese self-reliance. Chinese models are often praised for open weights and criticised for refusing or slanting answers on politically sensitive topics.

Famous chinese models

Further reading

imgHuawei PanGu 馃嚚馃嚦 ChinaHuawei PanGu Huawei's multimodal large model family for industry, from 1B to trillion-scaleMultimodal 路 Open weights 路 2023 Ernie Bot 馃嚚馃嚦 ChinaErnie Bot Baidu chatbot built on ERNIE language modelsAssistant 路 Open weights 路 2023 fnGLM 馃嚚馃嚦 ChinaGLM Open-weight Chinese model series used for chat and codingGeneral 路 Open weights 路 2021 fnDeepSeek-V4-Pro 馃嚚馃嚦 ChinaDeepSeek-V4-Pro 1.6T parameter MoE model with one million token contextGeneral 路 Open weights 路 2026 Qwen1.5 馃嚚馃嚦 ChinaQwen1.5 Open-weight language model series with nine sizes and strong multilingual support.General 路 Open weights 路 2024 MiMo-V2.5-Pro 馃嚚馃嚦 ChinaMiMo-V2.5-Pro Open-source MoE language model with 1M token contextGeneral 路 Open weights Qwen2 馃嚚馃嚦 ChinaQwen2 Open-weight language model series from AlibabaGeneral 路 Open weights 路 2024 Hunyuan-Large 馃嚚馃嚦 ChinaHunyuan-Large Open-source mixture of experts model with 389 billion parametersGeneral 路 Open weights 路 2024 DeepSeek-V2 馃嚚馃嚦 ChinaDeepSeek-V2 Efficient mixture-of-experts language model with 236 billion parametersGeneral 路 Open weights 路 2024 DeepSeek-V3 馃嚚馃嚦 ChinaDeepSeek-V3 Open-weight mixture of experts language model with 671 billion parameters.General 路 Open weights 路 2024 MiMo-V2-Flash 馃嚚馃嚦 ChinaMiMo-V2-Flash High-speed MoE language model for reasoning and agentsGeneral 路 Open weights fnGranite 4.0 馃嚭馃嚫 United StatesGranite 4.0 IBM's hybrid Mamba/transformer language models for enterprise workGeneral 路 Open weights 路 2025 DeepSeek-R1 馃嚚馃嚦 ChinaDeepSeek-R1 Open-weight reasoning model from ChinaReasoning 路 Open weights 路 2024

The weekly digest

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