The newest version of the Kimi model series by the Chinese AI lab Moonshot AI is anticipated to equal or potentially exceed the performance of Anthropic’s Opus 4.8, according to reports from unnamed sources.
The Kimi K2 models developed by Moonshot have received positive feedback in the open-source AI sector, achieving high scores on benchmark tests and displaying capabilities that closely follow the latest advanced models.
The upcoming Kimi K3 is expected to enhance this performance, further bridging the gap with proprietary models from companies like OpenAI and Anthropic. Reports indicate that Kimi K3 will become the largest open-weight AI model produced in China, boasting a parameter count between 2 trillion and 3 trillion, with a scheduled release in the near future.
Moonshot is also reportedly in the process of attracting new investments that could value the company at $31.5 billion, following a previous funding round in May that raised $2 billion at a $20 billion valuation.
This development arrives during an ongoing discussion about the merits of investing in closed-source AI models from firms such as OpenAI and Anthropic, which are often quite costly. Industry leaders express concerns that these AI labs might exploit the data provided by clients utilizing products like ChatGPT and Claude.
Executives are promoting their solutions as viable alternatives, urging businesses to consider more affordable open-source models from developers like DeepSeek, Z.ai, or Moonshot, and to customize them for their specific needs. This perspective is gaining traction as open-source models from Chinese companies increasingly catch up to their pricier, advanced counterparts.
