A New Chinese AI Model is Captivating Attention and Gaining Popularity

BEIJING — A new contender in the generative AI space, the Kimi K2 model from the Alibaba-backed startup Moonshot, has emerged as a low-cost alternative to OpenAI’s ChatGPT. Released late Friday night, Kimi K2 is an open-source large language model designed to provide coding capabilities at a significantly reduced price. This open-source approach, which allows free access to source code, contrasts with the strategies employed by major U.S. tech companies, except for some movements by Meta and Google. Meanwhile, OpenAI CEO Sam Altman announced an indefinite delay in the release of its first open-source model due to safety concerns, underscoring a competitive landscape where traditional giants may lag behind new entrants.

Kimi K2 has demonstrated strength in writing computer code, an area where many businesses see potential labor savings through generative AI. The model reportedly outperforms rivals, with claims of surpassing Anthropic’s Claude Opus 4 and achieving better overall performance than OpenAI’s GPT-4.1. In terms of cost, Kimi K2 offers a more affordable option for developers, charging only 15 cents per million input tokens and $2.50 for output tokens. This contrasts sharply with Claude Opus 4, which charges $15 per million input tokens, and GPT-4.1, which costs $2 for input and $8 for output.

Developers can use Kimi K2 freely, with minimal requirements for product disclosure. Initial reviews on social media have been largely positive, although some users noted issues with hallucinations, a common problem in generative AI. Industry analysts see Kimi K2 as a competitive model that is well-positioned in the rapidly evolving AI market. As companies like ByteDance, Tencent, and Baidu enhance their own offerings, the demand for Chinese alternatives to U.S. AI technologies continues to grow.