How much memory do I need to run Grok-1 (IQ1_S)?

The file size is 65.3 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 68.5 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.

Does Grok-1 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 Grok-1 with Ollama?

Run "ollama run hf.co/Arki05/Grok-1-GGUF:IQ1_S" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.

Arki05

Grok-1

Parameters: 316.5B
Max context length: 8,192
Architecture: grok
License: Commercial OK

License: apache-2.0

Memory needed to run IQ1_S at 8K tokens

68.5 GB Estimated

File size 65.3 GB Measured

VRAM it fits in (rough guide)

8 GB does not fit
12 GB does not fit
16 GB does not fit
24 GB does not fit
32 GB does not fit
48 GB does not fit

See devices and models for 64 GB

Macs cap how much memory the GPU can use, so leave headroom in unified memory.

Data as of Oct 6, 2026

About Grok-1

Grok-1 is a model published in GGUF format. GGUF files are published by Arki05/Grok-1-GGUF. It has about 316.5B parameters, the grok architecture, a maximum context length of 8,192 tokens (values written in the GGUF).

There are 21 quantizations, with file sizes from 65.3 GB (IQ1_S) to 335 GB (Q8_0) (actual sizes on Hugging Face).

Running IQ1_S (65.3 GB) with an 8,192-token context needs an estimated 68.5 GB of memory. For 8,192 tokens, a GPU with 64 GB or more of VRAM, or a Mac with 64 GB or more of unified memory, is a rough guide (macOS caps how much memory the GPU can use, so leave some headroom).

According to Arki05/Grok-1-GGUF, the license is Apache License 2.0, which allows commercial use (keep the license text and copyright notice when redistributing). 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

Arki05/Grok-1-GGUF

Published by the original developer About this label

Last checked: Oct 6, 2026

Architecture: grok; Supported before llama.cpp b4409
Parameters: 316.5B
Max context length: 8,192
License: apache-2.0 Commercial OK

File size and memory per quantization for Grok-1 (Arki05/Grok-1-GGUF)
Quantization	File size Measured	Quality guide	4,096	8,192 max
IQ1_S (9 parts)	65.3 GB	Very large	67.4 GB	68.5 GB
IQ1_M (9 parts)	72.3 GB	Very large	74.4 GB	75.5 GB
IQ2_XXS (9 parts)	83.9 GB	Very large	86.1 GB	87.2 GB
IQ2_XS (9 parts)	93.3 GB	Very large	95.4 GB	96.5 GB
IQ2_S (9 parts)	94.6 GB	Very large	96.7 GB	97.8 GB
IQ2_M (9 parts)	104 GB	Large	106 GB	107 GB
Q2_K_S (9 parts)	108 GB	Very large	110 GB	111 GB
Q2_K (9 parts) Q2_K	116 GB	Very large	118 GB	119 GB
Q2_K (9 parts) grok-1-Q2_K-split	116 GB	Very large	118 GB	119 GB
IQ3_XXS (9 parts)	122 GB	Large	124 GB	125 GB
IQ3_XS (9 parts) grok-1-IQ3_XS-split	130 GB	Large	132 GB	133 GB
IQ3_XS (9 parts) IQ3_XS	130 GB	Large	132 GB	133 GB
IQ3_S (9 parts)	137 GB	Medium	139 GB	140 GB
Q3_K_S (9 parts)	137 GB	Large	139 GB	140 GB
IQ3_M (9 parts)	144 GB	Medium	146 GB	147 GB
Q3_K_M (9 parts)	152 GB	Medium	154 GB	155 GB
Q3_K_L (9 parts)	163 GB	Medium	165 GB	166 GB
IQ4_XS (9 parts)	169 GB	Small	171 GB	172 GB
IQ4_NL (9 parts)	179 GB	Small	181 GB	182 GB
Q4_K (9 parts) Q4_K	192 GB	-	194 GB	195 GB
Q4_K (9 parts) grok-1-Q4_K-split	192 GB	-	194 GB	195 GB
Q5_K (9 parts)	224 GB	-	226 GB	228 GB
Q6_K (9 parts) Q6_K	259 GB	Very small	261 GB	262 GB
Q6_K (9 parts) grok-1-Q6_K-split	260 GB	Very small	262 GB	263 GB
Q8_0 (9 parts)	335 GB	Nearly identical	338 GB	339 GB

Ollama  ollama run hf.co/Arki05/Grok-1-GGUF:IQ1_S

llama.cpp  llama-server -hf Arki05/Grok-1-GGUF:IQ1_S

License and commercial use

Apache License 2.0 Commercial OK

Allows commercial use, modification and redistribution. When redistributing, keep the license text and copyright notice, and note changes in modified files.

The original model's license has not been fetched yet, so this is based on the GGUF repository's label. Original license text (checked Oct 18, 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

Example commands using IQ1_S. 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/Arki05/Grok-1-GGUF:IQ1_S

llama.cpp

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

llama-server -hf Arki05/Grok-1-GGUF:IQ1_S

LM Studio

Paste Arki05/Grok-1-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 Grok-1 (IQ1_S)?: The file size is 65.3 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 68.5 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can Grok-1 be used commercially?: The license is Apache License 2.0 (Commercial use allowed).Allows commercial use, modification and redistribution. When redistributing, keep the license text and copyright notice, and note changes in modified files. Check the original license text before deciding.
Does Grok-1 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 Grok-1 with Ollama?: Run "ollama run hf.co/Arki05/Grok-1-GGUF:IQ1_S" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.