Moonshot releases Kimi K3.1 Pro and says it cut its hacking skills

Moonshot releases Kimi K3.1 Pro and says it cut its hacking skills

Moonshot has released Kimi K3.1 Pro, an upgraded version of its open-weight Kimi K3, with full weights due by 4 November. It says the model nearly matches Claude Opus 5.5 on a coding test. Moonshot also says it removed hacking-related training data to limit misuse of the open release. Chinese AI lab Moonshot has released […] This story continues at The Next Web

Kimi K3.1 Pro Is Now The Strongest Model Out Of China: Cyber Data Stripped From Its Training But That Barely Holds Its Potential Back

With OpenAI and Anthropic both clamping down hard on the ability of China's AI labs to distill their leading models, many expected Moonshot's Kimi K3.1 Pro to flounder. And yet, not only has that eventually not come to pass, Moonshot seems to have surpassed the expectations of even its harshest critics, delivering a model that comfortably outranks both GPT-6 Astra and Claude Fable 5.1 on Artificial Analysis Intelligence Index. Moonshot's Kimi K3…

Chinese AI model Kimi K3.1 Pro is closing in on Anthropic's Claude Opus 5.5

Moonshot AI's Kimi K3.1 Pro scores 55 points on the Artificial Analysis Intelligence Index, an 11-point jump over its predecessor that puts it in fourth place. Only three Anthropic and Meta models rank higher. If Moonshot releases the weights as planned, the gap between open and proprietary models shrinks to just three points. The article Chinese AI model Kimi K3.1 Pro is closing in on Anthropic's Claude Opus 5.5 appeared first on The Decoder.

Moonshot AI releases Kimi K3.1 Pro open weight model to rival Claude

Moonshot AI built K3.1 Pro as a mixture-of-experts model with 2.8 trillion total parameters and 104 billion active at once, handling a 1 million token context window. The faster Kimi K3.1 Pro Highspeed variant is already available through Moonshot's apps and API.