LL

Llama 4 Scout 17B 16E Instruct

LlamaGGUF by bartowskiUpdated 2025-04-08
Download on Hugging Face
Parameters108B
Context10240K
Layers48
Architecturellama4
Smallest quant26.3 GB
Quick answer

Llama 4 Scout 17B 16E Instruct needs about 69.7 GB of memory at Q4_K_M (a 67.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 24 quants, from 26.3 GB (IQ1_M) to 215.6 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.

Llama 4 Scout 17B 16E Instruct – 110 B parameters, 10 M context, Mixture of Experts

Llama 4 Scout 17B 16E Instruct is a 110 B parameter model with a 10 M token context window. It is a Mixture of Experts architecture and is quantized by bartowski for use with llama.cpp. The model is available in multiple quantization levels, from IQ1_M to BF16.

Good at

  • Large context generation
  • High‑capacity instruction following
  • Mixture of Experts efficiency

Limits

  • No vision support
  • License is other – check usage terms
  • Requires large VRAM for full precision

Best for

  • Long‑form content creation
  • Complex instruction chains
  • Research prototyping

Who it’s for. Developers and researchers needing a large‑context LLM with Mixture of Experts. Suitable for projects that can accommodate the memory demands of a 110 B parameter model.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_M26.3 GB 28.4 GB
Too big Get file →
IQ2_XXS30.2 GB 32.3 GB
Too big Get file →
IQ2_XS32.9 GB 35.1 GB
Too big Get file →
IQ2_S34.3 GB 36.4 GB
Too big Get file →
IQ2_M37.1 GB 39.2 GB
Too big Get file →
Q2_K43.0 GB 45.1 GB
Too big Get file →
Q2_K_L44.0 GB 46.1 GB
Too big Get file →
IQ3_XXS45.0 GB 47.1 GB
Too big Get file →
IQ3_XS47.5 GB 49.6 GB
Too big Get file →
Q3_K_S49.8 GB 51.9 GB
Too big Get file →
IQ3_M50.3 GB (2 parts) 52.4 GB
Too big Get file →
Q3_K_M54.3 GB (2 parts) 56.4 GB
Too big Get file →
Q3_K_L57.8 GB (2 parts) 59.9 GB
Too big Get file →
Q3_K_XL58.7 GB (2 parts) 60.8 GB
Too big Get file →
IQ4_XS59.9 GB (2 parts) 62.0 GB
Too big Get file →
IQ4_NL63.0 GB (2 parts) 65.1 GB
Too big Get file →
Q4_063.1 GB (2 parts) 65.2 GB
Too big Get file →
Q4_K_M67.5 GB (2 parts) 69.7 GB
Too big Get file →
Q4_K_L68.3 GB (2 parts) 70.4 GB
Too big Get file →
Q4_169.1 GB (2 parts) 71.2 GB
Too big Get file →
Q5_K_L79.3 GB (2 parts) 81.4 GB
Too big Get file →
Q6_K_L89.3 GB (3 parts) 91.4 GB
Too big Get file →
Q8_0113.4 GB (3 parts) 115.5 GB
Too big Get file →
BF16215.6 GB (6 parts) 217.7 GB
Too big Get file →

For 8 GB VRAM, use IQ1_M (26.32 GB file, ~28.4 GB memory at 8 k context). For 16 GB VRAM, IQ2_S (34.34 GB file, ~36.4 GB memory). For 24 GB VRAM, Q3_K_M (54.32 GB file, ~56.4 GB memory).

What GPU do you need for Llama 4 Scout 17B 16E Instruct?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is IQ1_M (26.3 GB file, about 28.4 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_XXS (45.0 GB file, about 47.1 GB in use).

Which Llama 4 Scout 17B 16E Instruct quant should you choose?

Q4_K_M is the usual starting point for Llama 4 Scout 17B 16E Instruct: it balances file size and quality well. Here it is a 67.5 GB file.

The Q8_0 file is 113.4 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is IQ1_M at 26.3 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama4
Parameters108B (approximate, mixture of experts)
Layers48
Trained context10,485,760 tokens
KV cache (fp16)192 KB per token

Source

Quantized bybartowski
FamilyLlama
Licenseother
Downloads (HF)68,081
Last updated2025-04-08

Frequently asked questions about Llama 4 Scout 17B 16E Instruct

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

About 69.7 GB at Q4_K_M with an 8K context window. That is the 67.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Llama 4 Scout 17B 16E Instruct on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 28.4 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Llama 4 Scout 17B 16E Instruct GGUF?

It depends on the quant: IQ1_M is 26.3 GB, IQ2_XXS is 30.2 GB, IQ2_XS is 32.9 GB, IQ2_S is 34.3 GB, IQ2_M is 37.1 GB, Q2_K is 43.0 GB and 18 more.

What context length does Llama 4 Scout 17B 16E Instruct support?

The GGUF header lists a trained context length of 10,485,760 tokens (10240K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.

Where can I download Llama 4 Scout 17B 16E Instruct GGUF?

The files are on Hugging Face in the bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Llama 4 Scout 17B 16E Instruct use?

Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Llama 4 Scout 17B 16E Instruct on 8 GB VRAM?

Use the IQ1_M quantization; it requires about 28.4 GB of VRAM at 8 k context, so 8 GB is insufficient.

What is the context length of this model?

The model supports up to 10 M tokens of context.

Does this model support vision tasks?

No, the model has no vision capabilities.

Other GGUFs of Llama-4-Scout-17B-16E-Instruct

Similar models

Browse all →

More from bartowski

All by bartowski →

More Llama models

All Llama →

Other large models

All large models →