Kimi K3: 2.8 T‑parameter Mixture‑of‑Experts model with 1 M‑token context and native vision
MoonshotAI’s Kimi K3 pushes open LLMs to 2.8 trillion parameters, activating up to 104 billion weights per token via a 896‑expert MoE. It adds a 1‑million‑token context window and built‑in vision, delivering frontier‑level coding, reasoning, and agentic performance while remaining open‑source.