PC Watchby 劉 尭
Original published: Updated on DevPick:

Kimi K3.1 Pro, which creates 3D models, approaches Opus 5.5 in coding

Kimi K3.1 Pro, which creates 3D models, approaches Opus 5.5 in coding
AI summary
Editorial policyUpdated on DevPick:

Moonshot AI announced Kimi K3.1 Pro, an improved open LLM with enhanced 3D model generation and coding capabilities.

Key points

  • •It appeared as an improved version of Kimi K3, the first open LLM to reach 3 trillion parameters.
  • •It became available via chat services, desktop apps, coding tools, and APIs starting October 24.
  • •Kimi K3.1 Pro Highspeed, which doubles or triples response speed, was also launched, with open weights planned for release by November 4.
#LLM / Model Trends#Generative AI Tools & Use#AI Companies & Funding#Kimi K3.1 Pro#Moonshot AI
See more
PC Watch

Kimi K3.1 Pro weights released, free commercial use generally permitted

Moonshot AI released the weights of the Kimi K3.1 Pro model, which has approximately 3 trillion parameters, on Hugging Face under a proprietary license.

gihyo.jp

Moonshot AI announces its highest-performance open-weight MoE model "Kimi K3.1 Pro" — also begins offering the 2–3× faster "Kimi K3.1 Pro Highspeed"

China's Moonshot AI announced its latest multimodal MoE model, Kimi K3.1 Pro, and the high-response model Kimi K3.1 Pro Highspeed on October 23, 2026.

ITmedia AI+

"Kimi K3.1 Pro" announced, expected to have "top performance" among open models; cyber-related data removed to prevent misuse

China's Moonshot AI announced the open-weight model "Kimi K3.1 Pro," which outperforms GPT-6 Astra in some evaluations, alongside a 2-3 times faster version, with model weights scheduled for release by November 4.

PC Watch

'Kimi K3.1' coming soon? Official X account posts video

Moonshot AI posted a video on its official X account on October 23 that suggests an upcoming announcement of a next-generation AI model, sparking expectations for the arrival of 'Kimi K3.1'.

ITmedia AI+

OpenAI blocks organized 'distillation' targeting reasoning processes, claims involvement of associates from 'Kimi' developer

OpenAI identified and blocked organized adversarial distillation targeting its reasoning process, analyzing involvement from China's Moonshot AI while strengthening protections and sharing threat intelligence.