Llama 4 Scout 17B (16E)

109B MoE

Meta Llama 4

Meta Llama 4 Scout MoE (17B active / 109B total). Multimodal; needs ~68GB VRAM at Q4_K_M.

⬇ 29.3K HF downloads♥ 155 likesunsloth/Llama-4-Scout-17B-16E-Instruct-GGUF· stats from 6/25/2026
Pro GPU

10486K

Max Context

3

Quant Variants

GGUF Q4_K_M

Best Quality

97.6%

Accuracy Retained

Calculate VRAMHugging FaceCompare

Quantization Variants

Per-quant VRAM, quality loss, and inference speed on RTX 4090

FormatLevelBPWVRAMPPL LossSpeedActions
GGUFQ4_K_M4.8568.0 GB2.4%22 tok/s
CalcHF
GGUFQ3_K_M3.8755.0 GB4.8%26 tok/s
CalcHF
AWQINT4458.0 GB3.2%28 tok/s
CalcHF

Similar models

Compare with Llama 4
400B MoE

Llama 4 Maverick 17B (128E)

Meta Llama 4

Pro GPU
198.0 GBmin VRAM·97.8%accuracy

Llama 4 Maverick flagship MoE (17B active / 400B total). Multi-GPU or H100 cluster territory.

70B

Llama 3.1 70B Instruct

Meta Llama 3.1

Pro GPUMac / Apple Silicon
33.4 GBmin VRAM·98.8%accuracy

Meta's frontier 70B model. Requires 40GB+ VRAM; dual 3090 or M2 Ultra.

72B

Qwen2.5 72B Instruct

Alibaba Qwen2.5

Pro GPU
33.8 GBmin VRAM·98.9%accuracy

Flagship Qwen2.5. Requires dual 4090 or A100 80G. Exceptional reasoning at scale.

70B

Llama 3.3 70B Instruct

Meta Llama 3.3

Pro GPUMac / Apple Silicon
38.2 GBmin VRAM·99.0%accuracy

Latest Meta 70B with improved multilingual. Drop-in upgrade from Llama 3.1 70B.