- Moonshot AI has abruptly suspended new subscriptions to its powerful Kimi K3 artificial intelligence model after experiencing a rush of users.
- The startup claims that Kimi K3 achieves computational performance approaching that of the latest artificial intelligence models from OpenAI and Anthropic.
- Moonshot AI describes Kimi K3 as the world’s largest open-source AI system, with 2.8 trillion parameters.
- The surge in users has pushed the startup's computational capacity 'close to the limits'.
- China is making strides in the AI sector, with Moonshot and Alibaba unveiling models that they claim rival those of OpenAI and Anthropic.
- Moonshot's own testing ranks Kimi K3 consistently above nearly every US system, trailing only OpenAI’s GPT-5.6 Sol and Anthropic’s Claude Fable 5.
- Alibaba has also introduced a new model, Qwen3.8, which it claims is 'one of the most powerful models available today' and 'second only to Fable 5'.
- Both Moonshot and Alibaba are emphasizing a key difference from leading US labs by making their most advanced models publicly available.
- The emergence of these models raises questions about whether the vast sums of money US companies are investing in AI can secure a durable advantage.
Moonshot AI has suspended new subscriptions to its Kimi K3 model after a surge in users, claiming it rivals top-tier models from OpenAI and Anthropic. The model, with 2.8 trillion parameters, is touted as the world's largest open-source AI system.123568
The startup's rapid rise has pushed its computational capacity 'close to the limits,' prompting the halt in new subscriptions. Moonshot's internal testing ranks Kimi K3 above nearly every US system, trailing only OpenAI's GPT-5.6 Sol and Anthropic's Claude Fable 5, although it outperformed on certain benchmarks.4
In a competitive landscape, Alibaba has also entered the fray with its Qwen3.8 model, which it claims is 'one of the most powerful models available today.' Both companies emphasize their commitment to openness, contrasting with US firms that often keep advanced models behind closed doors.7
The implications of these developments are significant, as they challenge the notion that US companies can maintain a durable advantage in AI technology despite substantial investments in infrastructure and training. The emergence of these models raises questions about the future of AI development and competition on a global scale.
“The models also raise questions about whether the vast sums of money US companies are pouring into chips, data centers, and model training can secure a durable advantage,” analysts suggest.
Moonshot plans to release full model weights on July 27th, furthering its commitment to transparency in AI development.
“Moonshot describes Kimi K3 as the world's largest open-source AI system at 2.8 trillion parameters, with full model weights set to release on July 27. Meanwhile, Alibaba unveiled Qwen3.8, a 2.4 trillion parameter model it claims is "second only to Fable 5," as Chinese firms challenge US dominance with open-weight strategies.”
