Alibaba-backed startup Moonshot has launched its new language model Kimi K2, aimed at competing with well-known models from OpenAI and Anthropic.
Features of Kimi K2
Kimi K2 stands out for its coding capabilities, which are increasingly relevant as companies look to automate development processes. In an announcement posted on X and GitHub, Moonshot claimed that Kimi K2 outperformed Claude Opus 4 and surpassed OpenAI’s GPT-4.1 in several metrics. AI analyst Wei Sun from Counterpoint noted that the model is "globally competitive," and its open-source nature adds to its appeal.
Cost and Availability
Kimi K2 is available for free through an app and web browser. For commercial use, it costs 15 cents per million input tokens and $2.50 per million output tokens. In contrast, Claude Opus 4 charges $15 and $75, respectively. Moonshot states that Kimi K2 can be utilized by developers with minimal restrictions, aside from services with over 100 million users per month.
Moonshot's Market Prospects
Since its inception, Moonshot has established itself as a competitor to ChatGPT in China, while other companies like ByteDance and Tencent also vie for market share. Moonshot's emphasis on quality and accessibility has helped attract users. Moreover, other Chinese firms are also developing their AI products, signaling new trends in the AI landscape.
With the launch of Kimi K2, Moonshot is making a confident move on the international AI stage, offering a competitive and accessible product in an increasingly competitive environment.
