作者： Jakub Rusinowski · 最后更新： 2026年9月6日
Moonshot AI 2026年2月的突破之作。Kimi K2.5拥有1万亿总参数（32B活跃，384专家），创下HumanEval历史最高分99.0。在15万亿混合视觉和文本token上训练。具备独特的智能体群集架构。
| Licence | What it permits | Applies to |
|---|---|---|
Modified MIT | Commercial use permitted Commercial use permitted. No usage restrictions beyond attribution. | Kimi K2.5, Kimi K2.6, Kimi K2.7 Code |
Custom Open-Weight | Commercial use permitted Weights are downloadable and commercial use is permitted, subject to the licence’s acceptable-use terms. | Kimi K2.5 1T (32B Active) |
| Kimi K2.5 | Min 605 GB VRAM · Q4_K_M · 256,000 ctx · ollama run hf.co/moonshotai/Kimi-K2.5-Instruct-Q4_K_M |
| Kimi K2.6 | Min 605 GB VRAM · Q4_K_M · 256,000 ctx · ollama run hf.co/moonshotai/Kimi-K2.6-Instruct-Q4_K_M |
| Kimi K2.5 1T (32B Active) | Min 605 GB VRAM · Q4_K_M · 256,000 ctx · ollama run hf.co/moonshotai/Kimi-K2.5 |
| Kimi K2.7 Code | Min 605 GB VRAM · INT4 (native) · 262,144 ctx · |
Install Ollama then run: ollama run hf.co/moonshotai/Kimi-K2.5-Instruct-Q4_K_M
Minimum VRAM: 605 GB. For best results use Q4_K_M quantization.
Kimi K2.5 / K2.6 / K2.7 needs about 605 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Kimi K2.5 (605 GB, Q4_K_M); Kimi K2.6 (605 GB, Q4_K_M); Kimi K2.5 1T (32B Active) (605 GB, Q4_K_M); Kimi K2.7 Code (605 GB, INT4 (native)). On Apple Silicon, unified memory counts toward this requirement.
Kimi K2.5 / K2.6 / K2.7's smallest variant needs about 605 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 K2.5 / K2.6 / K2.7 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 hf.co/moonshotai/Kimi-K2.5-Instruct-Q4_K_M. This downloads Kimi K2.5 / K2.6 / K2.7 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.