Kimi Launches K3 Multimodal Model, Rivals GPT-5.6 Sol and Fable 5

Kimi is launching K3, a multimodal open-weight model with 2.8 trillion parameters and one million tokens of context. In the company's own benchmarks, it comes close to Claude Fable 5 and GPT 5.6 Sol while beating Opus 4.8 and GLM 5.2, in some cases by a wide margin. The model is also significantly pricier than its predecessor.

Full weights are scheduled for release by July 27. The K3 model's performance is a notable achievement for Kimi, a Chinese AI company. The company's benchmarks suggest that K3 is a strong contender in the multimodal model market.

However, the increased price point may be a barrier for some users. Kimi's move to a more expensive model may signal the end of super cheap Chinese AI. The company's decision to release the full weights of the model by July 27 could help developers and researchers build on its technology.

The K3 model's capabilities and pricing will likely have implications for businesses and consumers. As the AI market continues to evolve, Kimi's K3 model is poised to play a significant role. Why this matters: The launch of Kimi's K3 model has significant implications for the AI industry.

As a multimodal open-weight model, K3 has the potential to enable a wide range of applications, from natural language processing to computer vision. The model's performance, which rivals that of GPT-5.6 Sol and Fable 5, suggests that Kimi is a major player in the AI market. However, the increased price point may limit adoption among some users, particularly in China where cheap AI has been a hallmark of the industry.

As the market continues to evolve, it will be important to watch how Kimi's K3 model is used and whether its capabilities are sufficient to justify its cost. The release of the full weights by July 27 will also provide insight into how developers and researchers choose to build on Kimi's technology.