Written by Jakub Rusinowski · Last updated July 30, 2026
Moonshot AI's July 2026 flagship and its largest model to date — a 2.8-trillion-parameter Mixture-of-Experts that landed in the July 16–27 window. K3 roughly doubles the expert count of Kimi K2.5 (1T) while keeping active parameters modest (~48B per token), pushing Moonshot's lead on agentic coding, tool use, and long-horizon reasoning further ahead of K2.6. Natively multimodal with a 256K context. At this scale it is a data-center / API model — self-hosting needs well over a terabyte of memory even at Q4, so most users reach it through the Kimi API or the smaller K2.x community builds.
| Kimi K3 2.8T | Min 1400 GB VRAM · Q4_K_M · 256,000 ctx · |
Install Ollama then run: ollama run
Minimum VRAM: 1400 GB. For best results use Q4_K_M quantization.
Kimi K3 needs about 1400 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Kimi K3 2.8T (1400 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Kimi K3's smallest variant needs about 1400 GB, which exceeds a single RTX 4090 (24 GB). Use multiple GPUs, a higher-VRAM card, or Apple Silicon with large unified memory.
Q4_K_M is the best balance of quality and VRAM for Kimi K3 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
Install Ollama, then run: ollama run . This downloads Kimi K3 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.