Skip to content
localllm
日本語

bartowski

meta-llama_Llama-4-Scout-17B-16E-Instruct-old

Parameters
107.8B
Max context length
10,485,760
Architecture
llama4
License
Conditional
Base model
meta-llama/Llama-4-Scout-17B-16E-Instruct
License
llama4
llama.cpp
b5074+

About meta-llama_Llama-4-Scout-17B-16E-Instruct-old

meta-llama_Llama-4-Scout-17B-16E-Instruct-old is a GGUF model based on meta-llama/Llama-4-Scout-17B-16E-Instruct on Hugging Face. GGUF files are published by bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF. It has about 107.8B parameters, the llama4 architecture, a maximum context length of 10,485,760 tokens (values written in the GGUF).

There are 24 quantizations, with file sizes from 26.3 GB (IQ1_M) to 216 GB (BF16) (actual sizes on Hugging Face).

Running Q4_K_M (67.5 GB) with an 8,192-token context needs an estimated 70.1 GB of memory, and an estimated 71.4 GB at 32,768 tokens. For 8,192 tokens, a GPU with 96 GB or more of VRAM, or a Mac with 96 GB or more of unified memory, is a rough guide (macOS caps how much memory the GPU can use, so leave some headroom). The smallest, IQ1_M (26.3 GB), needs an estimated 28.9 GB at 8,192 tokens.

The original model is licensed under Llama 4 Community License, which allows commercial use with conditions such as scale or purpose. Check the original license for the terms of use.

File size and estimated memory per quantization

  • Up to 8 GB
  • Up to 16 GB
  • Up to 24 GB
  • Up to 48 GB
  • Up to 128 GB
  • More

File sizes are the actual values on Hugging Face. Memory is estimated for each context length (KV cache f16). The formula is in How memory is calculated.

All memory figures are Estimated (KV cache f16). It should run on a setup with at least this much VRAM (or Mac unified memory). Click a value to see the devices and models that fit that amount (Find by GPU / Mac).

About the "quality guide" (how far a quantization drifts from the original)

For each quantization type, how far the output drifts from the original model, grouped into six levels from values llama.cpp measured on Llama-3-8B.

The value is mean KLD (the difference between the original and quantized model in next-token probabilities); smaller means closer to the original. Quantizations made with an imatrix (importance data) come out slightly smaller. Levels use the value without an imatrix (or with one, for types that only have that).

These values were not measured on this model. Larger models and MoE models tend to drift less with the same type. Distributor-specific mixed types (such as UD-Q4_K_XL) cannot be matched by type name alone, so they show "-".

Quantization Quality guide KLD (no imatrix) KLD (imatrix) Size (Llama-3-8B)
Q8_0 Nearly identical 0.0014 - 7.96 GiB
Q6_K Very small 0.0055 - 6.14 GiB
Q5_K_M Very small 0.0108 - 5.33 GiB
Q5_K_S Small 0.0166 - 5.21 GiB
Q5_1 Small 0.0180 - 5.65 GiB
Q5_0 Small 0.0222 - 5.21 GiB
Q4_K_M Small 0.0313 0.0282 4.58 GiB
Q4_K_S Medium 0.0431 0.0320 4.37 GiB
IQ4_NL Small - 0.0357 4.35 GiB
IQ4_XS Small - 0.0363 4.14 GiB
Q4_1 Medium 0.0717 - 4.78 GiB
Q4_0 Medium 0.0719 - 4.34 GiB
Q3_K_L Medium 0.0872 0.0731 4.03 GiB
Q3_K_M Medium 0.1019 0.0844 3.74 GiB
IQ3_M Medium - 0.1025 3.53 GiB
IQ3_S Medium - 0.1113 3.42 GiB
IQ3_XS Large - 0.1387 3.28 GiB
IQ3_XXS Large - 0.1836 3.05 GiB
Q3_K_S Large 0.2282 0.1998 3.41 GiB
IQ2_M Large - 0.3260 2.74 GiB
Q2_K Very large 0.4451 0.3322 2.96 GiB
Q2_K_S Very large - 0.4034 2.96 GiB
IQ2_S Very large - 0.4392 2.56 GiB
IQ2_XS Very large - 0.5463 2.43 GiB
IQ2_XXS Very large - 0.8120 2.24 GiB
IQ1_M Very large - 1.3931 2.01 GiB
IQ1_S Very large - 2.2546 1.88 GiB

Source: llama.cpp (MIT), measured on Llama-3-8B against FP16, Wikitext-2, revision f364eb6f. Original table · Reading and choosing quantizations

GGUF source

bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF

Published by the original developer About this label

Using the original model requires a request or agreement on Hugging Face (this does not affect downloading the GGUF files).

Last checked: Nov 5, 2026

Architecture
llama4
llama.cpp b5074 and later
Parameters
107.8B
Max context length
10,485,760
License
llama4 Conditional
File size and memory per quantization for meta-llama_Llama-4-Scout-17B-16E-Instruct-old (bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF)
Quantization File size Measured Quality guide 4,096 8,192 16,384 32,768 65,536 131,072 262,144 10,485,760 max
IQ1_M 26.3 GB Very large 28.1 GB 28.9 GB 29.4 GB 30.2 GB 31.8 GB 35.0 GB 41.4 GB 544 GB
IQ2_XXS 30.2 GB Very large 31.9 GB 32.7 GB 33.2 GB 34.0 GB 35.6 GB 38.8 GB 45.3 GB 548 GB
IQ2_XS 32.9 GB Very large 34.7 GB 35.5 GB 36.0 GB 36.8 GB 38.4 GB 41.6 GB 48.1 GB 551 GB
IQ2_S 34.3 GB Very large 36.1 GB 36.9 GB 37.4 GB 38.2 GB 39.8 GB 43.0 GB 49.5 GB 552 GB
IQ2_M 37.1 GB Large 38.9 GB 39.7 GB 40.2 GB 41.0 GB 42.6 GB 45.8 GB 52.2 GB 555 GB
Q2_K 43.0 GB Very large 44.7 GB 45.5 GB 46.0 GB 46.8 GB 48.4 GB 51.7 GB 58.1 GB 561 GB
Q2_K_L 44.0 GB - 45.8 GB 46.6 GB 47.0 GB 47.8 GB 49.5 GB 52.7 GB 59.1 GB 562 GB
IQ3_XXS 45.0 GB Large 46.7 GB 47.5 GB 48.0 GB 48.8 GB 50.4 GB 53.6 GB 60.1 GB 563 GB
IQ3_XS 47.5 GB Large 49.2 GB 50.0 GB 50.5 GB 51.3 GB 52.9 GB 56.1 GB 62.6 GB 565 GB
Q3_K_S 49.8 GB Large 51.5 GB 52.3 GB 52.8 GB 53.6 GB 55.2 GB 58.4 GB 64.9 GB 567 GB
IQ3_M (2 parts) 50.3 GB Medium 52.1 GB 52.9 GB 53.4 GB 54.2 GB 55.8 GB 59.0 GB 65.4 GB 568 GB
Q3_K_M (2 parts) 54.3 GB Medium 56.1 GB 56.9 GB 57.4 GB 58.2 GB 59.8 GB 63.0 GB 69.4 GB 572 GB
Q3_K_L (2 parts) 57.8 GB Medium 59.6 GB 60.4 GB 60.8 GB 61.6 GB 63.3 GB 66.5 GB 72.9 GB 575 GB
Q3_K_XL (2 parts) 58.7 GB - 60.5 GB 61.3 GB 61.7 GB 62.5 GB 64.2 GB 67.4 GB 73.8 GB 576 GB
IQ4_XS (2 parts) 59.9 GB Small 61.6 GB 62.5 GB 62.9 GB 63.7 GB 65.3 GB 68.6 GB 75.0 GB 578 GB
IQ4_NL (2 parts) 63.0 GB Small 64.7 GB 65.6 GB 66.0 GB 66.8 GB 68.4 GB 71.7 GB 78.1 GB 581 GB
Q4_0 (2 parts) 63.1 GB Medium 64.8 GB 65.6 GB 66.1 GB 66.9 GB 68.5 GB 71.7 GB 78.2 GB 581 GB
Q4_K_M (2 parts) 67.5 GB Small 69.3 GB 70.1 GB 70.6 GB 71.4 GB 73.0 GB 76.2 GB 82.7 GB 585 GB
Q4_K_L (2 parts) 68.3 GB - 70.1 GB 70.9 GB 71.4 GB 72.2 GB 73.8 GB 77.0 GB 83.4 GB 586 GB
Q4_1 (2 parts) 69.1 GB Medium 70.9 GB 71.7 GB 72.1 GB 72.9 GB 74.6 GB 77.8 GB 84.2 GB 587 GB
Q5_K_L (2 parts) 79.3 GB - 81.1 GB 81.9 GB 82.4 GB 83.2 GB 84.8 GB 88.0 GB 94.4 GB 597 GB
Q6_K_L (3 parts) 89.3 GB - 91.0 GB 91.8 GB 92.3 GB 93.1 GB 94.7 GB 97.9 GB 104 GB 607 GB
Q8_0 (3 parts) 113 GB Nearly identical 115 GB 116 GB 116 GB 117 GB 119 GB 122 GB 129 GB 631 GB
BF16 (6 parts) 216 GB Original precision 217 GB 218 GB 219 GB 219 GB 221 GB 224 GB 231 GB 733 GB 
Ollama  ollama run hf.co/bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF:Q4_K_M
llama.cpp  llama-server -hf bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF:Q4_K_M

License and commercial use

Llama 4 Community License Conditional

Commercial use is allowed. If, on the Llama 4 release date, you or your affiliates had more than 700 million monthly active users, you need a separate license from Meta. When redistributing, or offering a product or service that contains the model, include the license text and prominently display "Built with Llama". If you use it to create, train or fine-tune an AI model that you distribute, start that model's name with "Llama". You must follow the Acceptable Use Policy.

Based on the license on the original model's Hugging Face page. Original license text (checked Nov 1, 2026)

The summary states facts this site read from the original text and is not legal advice. Always check the original license and its latest terms before use. All licenses

How to run

Mainline llama.cpp can load this model's format (llama4) from b5074 (Apr 7, 2025) onward. Ollama and LM Studio bundle llama.cpp internally, so an older app fails to load it with errors such as "unknown model architecture". If that happens, update the app.

Example commands using Q4_K_M. To use another quantization, replace the name with one from the table above. How to run local LLMs

Ollama

Run it by pointing directly at the GGUF on Hugging Face.

ollama run hf.co/bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF:Q4_K_M

llama.cpp

Pass the repository and quantization name to llama-server (or llama-cli) with -hf.

llama-server -hf bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF:Q4_K_M

LM Studio

Paste bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF into the search box, download it, and choose the quantization from the file list.

Frequently asked questions

How much memory do I need to run meta-llama_Llama-4-Scout-17B-16E-Instruct-old (Q4_K_M)?
The file size is 67.5 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 70.1 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can meta-llama_Llama-4-Scout-17B-16E-Instruct-old be used commercially?
The original model's license is Llama 4 Community License (Commercial use with conditions).Commercial use is allowed. If, on the Llama 4 release date, you or your affiliates had more than 700 million monthly active users, you need a separate license from Meta. When redistributing, or offering a product or service that contains the model, include the license text and prominently display "Built with Llama". If you use it to create, train or fine-tune an AI model that you distribute, start that model's name with "Llama". You must follow the Acceptable Use Policy. Check the original license text before deciding.
Does meta-llama_Llama-4-Scout-17B-16E-Instruct-old run in 8 GB, 12 GB, 16 GB or 24 GB of VRAM?
8 GB: no quantization fits; 12 GB: no quantization fits; 16 GB: no quantization fits; 24 GB: no quantization fits (estimated at 8,192 tokens with an f16 KV cache, files from the most-downloaded source).
What is the command to run meta-llama_Llama-4-Scout-17B-16E-Instruct-old with Ollama?
Run "ollama run hf.co/bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-old-GGUF:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.