Moonshot AI
2025-11-12 13:45 Zhejiang --- Moonshot AI’s Big Reveal: Kimi K2 Thinking Last week, Moonshot AI surprised the AI community by open-sourcing an enhanced version of Kimi K2 — dubbed Kimi K2 Thinking — under the slogan “Model as Agent”. This move sparked immediate excitement and global discussion. Hugging Face co-founder
Kimi K2
2025-11-11 — Beijing Update Last week, Moonshot AI released and open-sourced the enhanced version of Kimi K2 — Kimi K2 Thinking, positioned as a “model as Agent”, immediately generated strong buzz in the AI community. Hugging Face co-founder Thomas Wolf wondered aloud: > "Is this another DeepSeek moment?" --- First
AI hardware
# Qualcomm Announces AI Chips to Compete with AMD and Nvidia — Stock Soars 11% [](https://substackcdn.com/image/fetch/$s_!tkew!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2773e195-7d24-4120-a6dd-031adf1542e8_
AI Models
Open Source vs. Closed Source Models: The Real Showdown Overview After a four-month hiatus, Kimi has launched the inference-ready version of its flagship open-source model K2 — now renamed K2 Thinking. It is the most powerful open-source thinking model Kimi has ever built, equipped with: * 1 trillion parameters * Mixture of Experts
Kimi K2
Kimi K2 Thinking — Overview Kimi K2 Thinking is a new release from Moonshot, a Chinese AI lab, building on their earlier Kimi K2 model introduced in July. This Thinking variant maintains 1 trillion parameters (Mixture-of-Experts, 32B active) and is released under a custom-modified MIT license — meaning it is not fully
RTX 4090
# Fine-Tuning Ultra-Large Parameter Models — Now Possible on Consumer GPUs Fine-tuning extremely large models has undergone a **dramatic transformation**: You now only need **2–4 consumer-grade GPUs** (such as the RTX 4090) to locally fine-tune models as huge as **DeepSeek 671B** or **Kimi K2 1TB**. 
Chinese AI
Chinese AI Models Winning Over Global Enterprises While OpenAI has been promoting ChatGPT extensively, Chinese large language models are capturing international markets by sheer capability. --- Airbnb Chooses Alibaba’s Qwen Over OpenAI Recently, Airbnb Co‑founder and CEO Brian Chesky publicly praised Alibaba’s Qwen model: > We rely
AI shift
 # Silicon Valley Billionaire Votes with His Feet — A Signal from the Heart of Tech An unexpected and powerful signal is emerging right from the core of **Silicon Valley**. On the influential *All-In* tech investment podcast, **Chamath Palihapitiya** revealed