Not Just DeepSeek, Several Other Chinese AI Firms Rise To Significance With Low Cost Models

Alibaba’s recent announcement about its partnership with Apple to support artificial intelligence services for iPhones in China has once again drawn attention to the rapidly evolving AI landscape in the country. This move follows the developments surrounding DeepSeek, highlighting China’s progress in AI technology.

Here are some of the leading AI services and their developers making waves in China:

Doubao: ByteDance’s AI Powerhouse

Developed by ByteDance, Doubao has quickly become the most widely used AI chatbot in China. Despite entering the large language model (LLM) market later than others, ByteDance has established itself as a major player.

The company’s latest model, Doubao-1.5-pro, launched in January, has been reported to surpass GPT-4 in certain performance metrics. ByteDance is also known for offering cost-effective LLM solutions. According to a Barclays report, Doubao-1.5 is priced at just 3% to 4% of GPT-4’s cost, making it the most affordable AI offering in China.

ByteDance attributes this pricing advantage to the “Mixture of Experts” (MoE) framework, which is also utilized by DeepSeek and other AI models in China.

Qwen: Alibaba’s Flagship AI Model

Qwen, Alibaba’s primary LLM, was launched in March 2023 and has been recognized among top-tier AI models in global benchmarks. In January, Alibaba introduced Qwen 2.5-Max, claiming that it outperforms the well-regarded DeepSeek-V3.

A strong supporter of open-source AI, Alibaba announced in September that it would make many of its AI models publicly available, reflecting its commitment to advancing AI development.

Ernie: Baidu’s Advanced AI Model

Baidu was one of the first Chinese companies to launch a ChatGPT-like AI chatbot, Ernie Bot. The company asserts that its latest model, Ernie 4.0, matches GPT-4 in capability.

On February 14, Baidu revealed plans to open-source Ernie 4.5 starting from June 30, a major strategic shift as AI competition in China intensifies.

Hunyuan: Tencent’s AI Ambitions

Tencent introduced its flagship model, Hunyuan, in September, followed by its chatbot, Yuanbao, in May 2024. The company’s vast user base, particularly through popular platforms like WeChat, gives it an edge in reaching consumers.

Tencent also supports open-source AI development and focuses on multi-modal capabilities, including text, image, and video generation.

GLM: Zhipu AI’s Contribution to AI Innovation

Zhipu AI developed GLM, and its latest model, GLM4, is reported to outperform GPT-4 in certain benchmarks.

Originally developed in a Tsinghua University laboratory, Zhipu is backed by major investors, including Alibaba, Tencent, and Zhongguancun Science City Innovation Development. However, in January, the U.S. Commerce Department added Zhipu to its export control entity list.

Kimi: Moonshot’s AI Breakthrough

Moonshot, backed by major investors such as Alibaba, Tencent, and Hongshan, developed Kimi, a chatbot recognized for its long-context processing abilities. The company claims that its model can process up to two million input tokens per query.

Just two days before DeepSeek released the R1 model, Moonshot launched its own reasoning model, Kimi 1.5, on January 20.

MiniMax: A Pioneer in MoE Framework

MiniMax has been a leader in researching the MoE framework and open-sourced its most advanced model, MiniMax-01, in January.

Founded in December 2021, MiniMax predates the AI boom sparked by OpenAI’s ChatGPT. It is backed by investors such as Alibaba, Hongshan, Gaorong Capital, and IDG. Apart from LLMs, MiniMax has also gained popularity with consumer applications like Takie and Xingye.

01.AI: Cost-Efficient AI Development

01.AI follows an open-source model similar to DeepSeek and is a pioneer in the MoE framework. The company emphasizes cost efficiency, claiming its model operates at 1/40 the cost of GPT-4 inference.

Founded by Kai-Fu Lee, a former head of Google China, 01.AI has formed a strategic partnership with Alibaba to establish a research lab focused on LLM development.

Baichuan: Early Leader in Open-Source AI

Baichuan, founded in April 2023 by former executives of search engine Sogou, made headlines in June 2023 when it became the first Chinese company to launch an open-source AI model. The company is led by Wang Xiaochuan, former CEO of Sogou.

With major players like Alibaba, Tencent, ByteDance, and Baidu investing heavily in AI, China’s artificial intelligence industry is evolving rapidly. From cost-effective models to open-source initiatives, these companies are shaping the future of AI, making China a formidable competitor in the global AI race.

ALSO READ: Samsung Galaxy F06 5G Debuts In India: 50MP Camera & 5000mAh Battery At An Unbeatable Price