RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can I run Llama 4 Scout 17B 16E Instruct locally?

Llama 4 Scout 17B 16E Instruct by meta-llama has 109B parameters (18B active per token, 16 experts, 1 used per token). At 4-bit with an 8K context it needs about 63.2 GB of GPU memory. It fits on 22 of 65 GPUs and Macs we track.

63.2 GBQ4_K_M · 8K context
nonesmallest consumer GPU that fits
M2 Max 96GBsmallest Mac that fits
10Mmax context
GQAattention type

Memory needed by quant and context length

GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use real file sizes from unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF where available (marked •), otherwise parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx32K ctx128K ctx
2-bit (Q2_K) •rough38.4 GB39.1 GB43.6 GB61.6 GB
3-bit (Q3_K_M) •usable49.7 GB50.5 GB55.0 GB73.0 GB
4-bit (Q4_K_M) •good, most popular62.4 GB63.2 GB67.7 GB85.7 GB
5-bit (Q5_K_M) •very good72.8 GB73.6 GB78.1 GB96.1 GB
6-bit (Q6_K) •near lossless83.9 GB84.6 GB89.1 GB107 GB
8-bit (Q8_0) •effectively lossless108 GB109 GB113 GB131 GB
16-bit (F16/BF16) •original weights202 GB203 GB208 GB226 GB

Which GPUs and Macs run Llama 4 Scout 17B 16E Instruct?

Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).

HardwareMemoryFits?Best quantNeedsMax tok/s (est.)
RTX 3060 12GB12 GBNo———details
RTX 4060 Ti 16GB16 GBNo———details
RTX 309024 GBNo———details
RTX 409024 GBNo———details
RTX 509032 GBNo———details
RTX 5070 Ti16 GBNo———details
RX 7900 XTX24 GBNo———details
Mac M4 Pro 48GB48 GBNo———details
Mac M4 Max 128GB128 GBFitsQ6_K84.6 GB37details
Mac M3 Ultra 512GB512 GBFitsF16203 GB23details
DGX Spark128 GBFitsQ6_K84.6 GB19details
RTX 407012 GBNo———details
RTX 5060 Ti 16GB16 GBNo———details
RTX 40608 GBNo———details
Arc B58012 GBNo———details
RTX 3060 Ti8 GBNo———details
RTX 30708 GBNo———details
RTX 3080 10GB10 GBNo———details
RTX 3080 12GB12 GBNo———details
RTX 3090 Ti24 GBNo———details
RTX 4060 Ti 8GB8 GBNo———details
RTX 4070 Super12 GBNo———details
RTX 4070 Ti12 GBNo———details
RTX 4070 Ti Super16 GBNo———details
RTX 408016 GBNo———details
RTX 4080 Super16 GBNo———details
RTX 50608 GBNo———details
RTX 507012 GBNo———details
RTX 508016 GBNo———details
2x RTX 30902×24 GBFitsQ2_K39.1 GB142details
2x RTX 40902×24 GBFitsQ2_K39.1 GB153details
2x RTX 50902×32 GBFitsQ3_K_M50.5 GB208details
RTX A600048 GBFitsQ2_K39.1 GB117details
RTX 6000 Ada48 GBFitsQ2_K39.1 GB146details
RTX PRO 6000 Blackwell96 GBFitsQ5_K_M73.6 GB141details
L40S48 GBFitsQ2_K39.1 GB131details
A100 40GB40 GBNo———details
A100 80GB80 GBFitsQ4_K_M63.2 GB188details
H100 PCIe80 GBFitsQ4_K_M63.2 GB184details
H100 SXM80 GBFitsQ4_K_M63.2 GB309details
H200141 GBFitsQ8_0109 GB252details
RX 76008 GBNo———details
RX 7700 XT12 GBNo———details
RX 7800 XT16 GBNo———details
RX 7900 GRE16 GBNo———details
RX 7900 XT20 GBNo———details
RX 907016 GBNo———details
RX 9070 XT16 GBNo———details
Ryzen AI Max+ 395 128GB128 GBFitsQ6_K84.6 GB17details
Arc A770 16GB16 GBNo———details
Mac M1 16GB16 GBNo———details
Mac M2 24GB24 GBNo———details
Mac M3 24GB24 GBNo———details
Mac M4 16GB16 GBNo———details
Mac M4 32GB32 GBNo———details
Mac M1 Pro 32GB32 GBNo———details
Mac M3 Pro 36GB36 GBNo———details
Mac M4 Pro 24GB24 GBNo———details
Mac M4 Pro 64GB64 GBFitsQ2_K39.1 GB42details
Mac M1 Max 64GB64 GBFitsQ2_K39.1 GB61details
Mac M2 Max 96GB96 GBFitsQ4_K_M63.2 GB37details
Mac M3 Max 128GB128 GBFitsQ6_K84.6 GB27details
Mac M4 Max 64GB64 GBFitsQ2_K39.1 GB83details
Mac M2 Ultra 192GB192 GBFitsQ8_0109 GB42details
Mac M3 Ultra 256GB256 GBFitsQ8_0109 GB43details

Prefer an API?

Meta: Llama 4 Scout is on OpenRouter at $0.10 per 1M input tokens and $0.30 per 1M output tokens. Price details and history →

Questions

How much VRAM does Llama 4 Scout 17B 16E Instruct need?

About 63.2 GB at 4-bit (Q4_K_M) with an 8K context: 60.9 GB of weights (real GGUF file size), 1.5 GB of KV cache and 0.8 GB of runtime overhead. At 8-bit it needs about 109 GB.

What is the smallest GPU that runs Llama 4 Scout 17B 16E Instruct?

No single consumer GPU in our table fits it at Q4_K_M with 8K context; the smallest option that does is the H100 PCIe (80 GB).

Can I run Llama 4 Scout 17B 16E Instruct on a Mac?

Yes. At Q4_K_M with 8K context it fits on a Mac M2 Max 96GB or larger, using the default macOS GPU memory limit.

Similar models

Same family or a similar parameter count, with the VRAM each needs at Q4_K_M and 8K context.

Llama 4 Scout 17B 16E

109B · 64.1 GB at 4-bit

Llama 3.3 70B Instruct

71B · 43.3 GB at 4-bit

Llama 3.1 70B

71B · 42.7 GB at 4-bit

GLM 4.5V

108B · 61.4 GB at 4-bit

GLM 4.6V

108B · 67.8 GB at 4-bit

GLM 4.5 Air

110B · 70.1 GB at 4-bit

Model facts

Hugging Face repometa-llama/Llama-4-Scout-17B-16E-Instruct
Parameters108,641,793,536 (109B) · ~18B active (computed from config)
Layers / hidden size48 / 5120
AttentionGQA; 40 heads, 8 KV heads, head dim 128
KV cache per token (FP16)192.0 KiB
Max context10,485,760 tokens
Licenseother (gated: accept terms on Hugging Face)
Released on HFApr 2, 2025
Downloads (30d) / likes110,352 / 1,391

Sources: Hugging Face model API · config.json (unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.