Models › Nanbeige4.2-3B

Nanbeige4.2-3B: VRAM needs and speed by GPU

At Q4_K_M, Nanbeige4.2-3B needs about 3.6 GB of VRAM. No measured runs yet. Speeds below are estimates. Submit a run.

4.17B parameters · license: apache-2.0 · 262,144-token context · GGUF published Jul 28, 2026 · GGUF files on Hugging Face

Memory needed per quantization

QuantFile sizeNeeds (estimate)Smallest card it fits
Q2_K1.8 GB2.7 GB8 GB cards
IQ2_M1.9 GB2.8 GB8 GB cards
IQ3_XXS1.9 GB2.8 GB8 GB cards
IQ3_XS2.0 GB2.9 GB8 GB cards
Q3_K_S2.1 GB3.0 GB8 GB cards
IQ3_M2.2 GB3.1 GB8 GB cards
Q3_K_M2.2 GB3.1 GB8 GB cards
Q3_K_L2.3 GB3.2 GB8 GB cards
Q2_K_L2.3 GB3.2 GB8 GB cards
IQ4_XS2.4 GB3.3 GB8 GB cards
Q4_02.5 GB3.4 GB8 GB cards
IQ4_NL2.5 GB3.4 GB8 GB cards
Q4_K_S2.6 GB3.5 GB8 GB cards
Q4_K_M2.7 GB3.6 GB8 GB cards
Q4_12.7 GB3.6 GB8 GB cards
Q3_K_XL2.8 GB3.7 GB8 GB cards
Q5_K_S3.0 GB3.9 GB8 GB cards
Q4_K_L3.1 GB4.0 GB8 GB cards
Q5_K_M3.1 GB4.0 GB8 GB cards
Q5_K_L3.4 GB4.3 GB8 GB cards
Q6_K3.6 GB4.5 GB8 GB cards
Q6_K_L3.8 GB4.7 GB8 GB cards
Q8_04.4 GB5.3 GB8 GB cards
BF168.3 GB9.2 GB12 GB cards

Expected speed on popular hardware

HardwareBest quant that fitsExpected tok/s
NVIDIA RTX 3060Q8_069.5 estimate
NVIDIA RTX 4060 TiQ8_055.2 estimate
NVIDIA RTX 3090Q8_0148 estimate
NVIDIA RTX 4090Q8_0164 estimate
NVIDIA RTX 5090Q8_0259 estimate
AMD Radeon RX 7900 XTXQ8_0165 estimate
Apple M4 Pro (20-core GPU) (64 GB)Q8_052.2 estimate
Apple M3 Max (40-core GPU) (64 GB)Q8_076.2 estimate
AMD Ryzen AI Max+ 395 (Radeon 8060S)Q8_046.9 estimate

Check your own card

Too big for your card?

Run Nanbeige4.2-3B 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.