RigBench data 2026-10-22

Home › Models › Nex-N2-Pro

Nex-N2-Pro: VRAM needs and speed by GPU

At Q4_K_M, Nex-N2-Pro needs about 244.8 GB of VRAM. We have 1 measured speed results for it.

Ranked #18 among open-weight models by the Artificial Analysis Intelligence Index (provisional), checked 2026-10-22T18:00:53Z. All leading open models.

396.35B parameters (mixture-of-experts, ~17.0B active) · license: apache-2.0 · 262,144-token context · GGUF published 2026-06-10T15:26:21.000Z · GGUF files on Hugging Face

Measured speeds

ChipGeneration tok/sPrompt tok/sQuantBackendTestSource
NVIDIA DGX Spark (GB10)18.9–IQ1_MCUDA decode per session, llama-server, 1 concurrent request, 1024-token promptjvr0x/dgx-spark-bench results details

Memory needed per quantization

QuantFile sizeNeeds (estimate)Smallest card it fits
IQ1_S81.8 GB84.8 GB96 GB cards
IQ1_M91.4 GB94.4 GB96 GB cards
IQ2_XXS106.3 GB109.3 GBmulti-GPU / large unified memory
IQ2_XS118.5 GB121.5 GBmulti-GPU / large unified memory
IQ2_S120.5 GB123.5 GBmulti-GPU / large unified memory
IQ2_M133.1 GB136.1 GBmulti-GPU / large unified memory
Q2_K139.5 GB142.5 GBmulti-GPU / large unified memory
Q2_K_L140.5 GB143.5 GBmulti-GPU / large unified memory
IQ3_XXS166.1 GB169.1 GBmulti-GPU / large unified memory
Q3_K_S172.9 GB175.9 GBmulti-GPU / large unified memory
IQ3_XS181.4 GB184.4 GBmulti-GPU / large unified memory
Q3_K_M181.5 GB184.5 GBmulti-GPU / large unified memory
Q3_K_L189.5 GB192.5 GBmulti-GPU / large unified memory
IQ3_M189.6 GB192.6 GBmulti-GPU / large unified memory
Q3_K_XL190.4 GB193.4 GBmulti-GPU / large unified memory
IQ4_XS212.2 GB215.2 GBmulti-GPU / large unified memory
IQ4_NL224.5 GB227.5 GBmulti-GPU / large unified memory
Q4_0225.4 GB228.4 GBmulti-GPU / large unified memory
Q4_K_S232.8 GB235.8 GBmulti-GPU / large unified memory
Q4_K_M241.8 GB244.8 GBmulti-GPU / large unified memory
Q4_K_L242.6 GB245.6 GBmulti-GPU / large unified memory
Q4_1249.1 GB252.1 GBmulti-GPU / large unified memory
Q5_K_S274.0 GB277.0 GBmulti-GPU / large unified memory
Q5_K_M283.3 GB286.3 GBmulti-GPU / large unified memory
Q6_K342.4 GB345.4 GBmulti-GPU / large unified memory
Q8_0421.5 GB424.5 GBmulti-GPU / large unified memory

Expected speed on popular hardware

HardwareBest quant that fitsExpected tok/s
NVIDIA RTX 3060Does not fit
NVIDIA RTX 4060 TiDoes not fit
NVIDIA RTX 3090Does not fit
NVIDIA RTX 4090Does not fit
NVIDIA RTX 5090Does not fit
AMD Radeon RX 7900 XTXDoes not fit
Apple M4 Pro (20-core GPU)Does not fit
Apple M3 Max (40-core GPU)Does not fit
AMD Ryzen AI Max+ 395 (Radeon 8060S)IQ1_M26.6 estimate

Check your own card

Too big for your card?

Run Nex-N2-Pro on a rented 24-80 GB GPU by the hour.

Rent a bigger GPU on RunPod

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.