Model Spend Arena2ND ED.
397B · dense · 2026-08-14

Nex-N2-Pro

Runs on Data centre · Hugging Face ↗

vLLMSGLang

Coding index 59.1 (Artificial Analysis)

Nex AGI's 397B MoE, open-weight, a strong non-frontier open coder.

Sizes on disk

Real GGUF file sizes = weight VRAM.

QuantizationSize
Q3_K_M181.5 GB
IQ4_XS212.2 GB
Q4_0225.4 GB
Q4_K_S232.8 GB
Q4_K_M241.8 GB
Q4242.6 GB
Q4_1249.1 GB
Q2_K280.0 GB
Q5_K_M283.3 GB
Q6_K342.4 GB
Q8_0421.5 GB

Config tips

~242 GB at Q4 — multi-node. Serve with vLLM / SGLang. Origin of the lab is not well documented, but the weights and the AA score are real.