Llama 4 Scout (109B / 17B-16E MoE) — VRAM Requirements, Quantization Matrix & GPU Guide
Llama 4 Scout (109B / 17B-16E MoE) (Meta Llama, Llama 4 Community) is a Mixture-of-Experts (MoE) architecture with 109B resident parameters (17.2B active per token), 48 layers, and Hybrid Sliding-Window (SWA) supporting up to 10,485,760 context tokens.
Interactive Llama 4 Scout (109B / 17B-16E MoE) Memory & Context Simulator
Precision & Quantization VRAM Matrix (8K / 32K / 128K Context)
| Precision / Format | Effective bpw | Weights Only | Total @ 8K ctx | Total @ 32K ctx | Total @ 128K ctx | Recommended Tier (@ 32K) |
|---|---|---|---|---|---|---|
| FP16 / BF16 | 16.00 bpw | 203.03 GB | 211.82 GB | 212.95 GB | 217.45 GB | Mac Studio 512GB |
| FP8 | 8.50 bpw | 107.86 GB | 113.80 GB | 114.92 GB | 119.42 GB | NVIDIA H200 141GB |
| Q8_0 | 8.50 bpw | 107.86 GB | 113.80 GB | 114.92 GB | 119.42 GB | NVIDIA H200 141GB |
| Q6_K | 6.56 bpw | 83.24 GB | 88.44 GB | 89.57 GB | 94.07 GB | RTX PRO 6000 96GB |
| Q5_K_M | 5.69 bpw | 72.20 GB | 77.07 GB | 78.20 GB | 82.70 GB | RTX PRO 6000 96GB |
| Q4_K_M | 4.85 bpw | 61.54 GB | 66.09 GB | 67.22 GB | 71.72 GB | NVIDIA A100 80GB |
| AWQ INT4 | 4.50 bpw | 57.10 GB | 61.52 GB | 62.64 GB | 67.14 GB | NVIDIA A100 80GB |
| IQ4_XS | 4.25 bpw | 53.93 GB | 58.25 GB | 59.37 GB | 63.87 GB | NVIDIA A100 80GB |
Hardware Compatibility & Decode Speed (Llama 4 Scout (109B / 17B-16E MoE) @ Q4_K_M, 32K Context)
| Hardware Tier | Memory & Bandwidth | Required (32K) | Fit Verdict | Est. Decode Speed |
|---|---|---|---|---|
| RTX 3090 24GB | 24 GB · 936 GB/s | 67.22 GB (280%) | Out of Memory (OOM) | — |
| RTX 4090 24GB | 24 GB · 1008 GB/s | 67.22 GB (280%) | Out of Memory (OOM) | — |
| RTX 5090 32GB | 32 GB · 1792 GB/s | 67.22 GB (210%) | Out of Memory (OOM) | — |
| Apple M-Max 64GB | 64 GB · 546 GB/s | 65.95 GB (103%) | Out of Memory (OOM) | — |
| Mac Studio 128GB | 128 GB · 546 GB/s | 65.95 GB (52%) | Fits Comfortably | 24.3 tok/s |
| Mac Studio 192GB | 192 GB · 800 GB/s | 65.95 GB (34%) | Fits Comfortably | 35.6 tok/s |
| NVIDIA A100 80GB | 80 GB · 2039 GB/s | 67.22 GB (84%) | Tight Fit (>90% Usable) | 102.0 tok/s |
| NVIDIA H100 80GB | 80 GB · 3350 GB/s | 67.22 GB (84%) | Tight Fit (>90% Usable) | 167.5 tok/s |
Verified Transformer Architecture Parameters
- Checkpoint ID
- meta-llama/Llama-4-Scout-17B-16E-Instruct
- Ollama Tag
- llama4:scout
- License / Release
- Llama 4 Community · 2025-Q2
- Topology & Attention
- Mixture-of-Experts (MoE)
- Attention Mechanism
- Hybrid Sliding-Window (SWA)
- Total Params (B)
- 109B
- Active Params / Token (B)
- 17.2B
- Hidden Layers / Hidden Size
- 48 / 5120
- Query Heads / KV Heads
- 40Q / 8KV (d=128)
- Sliding Window
- 8192 (75% local)
- MoE Routed Experts
- 1 active / 16 total
- Vocab / Max Position
- 202,048 / 10,485,760
Quantized formats store per-block scale factors and zero-points alongside packed integer weights. A 4-bit GGUF Q4_K_M checkpoint averages 4.85 effective bits per parameter (not 4.00), while Q8_0 averages 8.50 bpw. For Mixture-of-Experts models, all routed and shared experts remain resident in memory to prevent token-by-token host paging.
Standard Grouped-Query Attention allocates 2 × L × H_kv × d_head × S × B × bytes_kv across layers. DeepSeek Multi-head Latent Attention (MLA) compresses keys and values into L × (d_kv_lora + d_qk_rope) × S × B × bytes_kv, slashing long-context memory by over 90%. Hybrid sliding-window models cap local-attention layers at min(S, window_size) while scaling global layers linearly with S.
Quick-Start vLLM & Ollama Commands for Llama 4 Scout (109B / 17B-16E MoE)
vllm serve meta-llama/Llama-4-Scout-17B-16E-Instruct \
--tensor-parallel-size 1 \
--max-model-len 32768 \
--max-num-seqs 8 \
--gpu-memory-utilization 0.95 \
--kv-cache-dtype fp8 \
--enable-prefix-caching \
--port 8000# Save as Modelfile, then run: ollama create llama-4-scout-custom -f Modelfile
FROM llama4:scout
PARAMETER num_ctx 32768
PARAMETER num_batch 512
PARAMETER flash_attn true
# Environment variables for ollama serve:
# OLLAMA_NUM_PARALLEL=8 OLLAMA_FLASH_ATTENTION=1 OLLAMA_KV_CACHE_TYPE=q8_0 ollama serveCompare Sibling & Similar-Scale Open-Weight Models
Llama 3.3 70B Instruct
Meta Llama · 70.6B (GQA)
Llama 3.1 8B Instruct
Meta Llama · 8.03B (GQA)
Llama 4 Maverick (400B / 17B-128E MoE)
Meta Llama · 402B (GQA)
Llama 3.1 405B Instruct
Meta Llama · 405B (GQA)
Cohere Command A (111B)
Cohere · 111B (HYBRID_SWA)
Mistral Large 2411 (123B)
Mistral · 123B (GQA)