Skip to content

moonshotai_Kimi-K2-Instruct-mlx-3bit

mlx-community/moonshotai_Kimi-K2-Instruct-mlx-3bit
3-bit Open on Hugging Face
Parameters
1.0T
Weights (est.)
≈ 449 GB
Mac memory (est.)
512 GB+

Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.

Run it

With mlx-lm on an Apple silicon Mac:

pip install mlx-lmmlx_lm.generate --model mlx-community/moonshotai_Kimi-K2-Instruct-mlx-3bit --prompt "Hello"

Details

Base model
moonshotai/Kimi-K2-Instruct (quantized) · Kimi K2 Instruct on the radar →
Vendor
Moonshot · Kimi
Task
Text generation
Quantisation
3-bit
Licence
modified-mit
Downloads
219 in the last 30 days · 4.5K all time · 2 likes
Published
2025-08-22 · updated 2025-08-23 (1y ago)
Tags
kimi_k2, custom_code

Other MLX builds of moonshotai/Kimi-K2-Instruct

Build Quantisation Weights Mac memory Downloads
moonshotai_Kimi-K2-Instruct-mlx-3bit (this one) 3-bit ≈ 449 GB 512 GB+ 219
Kimi-K2-Instruct-4bit 4-bit ≈ 578 GB > 512 GB 1.2K