How much memory do I need to run Mistral-Small-Instruct-2409 (Q4_K_M)?

The file size is 13.3 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 15.8 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.

Does Mistral-Small-Instruct-2409 run in 8 GB, 12 GB, 16 GB or 24 GB of VRAM?

8 GB: up to IQ1_M (est. 7.75 GB); 12 GB: up to Q3_K_S (est. 12.1 GB); 16 GB: up to Q4_K_M (est. 15.8 GB); 24 GB: up to Q6_K (est. 20.7 GB) (estimated at 8,192 tokens with an f16 KV cache, files from the most-downloaded source).

What is the command to run Mistral-Small-Instruct-2409 with Ollama?

Run "ollama run hf.co/MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.

mistralai

Mistral-Small-Instruct-2409

Parameters: 22.2B
Max context length: 32,768
Architecture: llama
License: Non-commercial

Base model: mistralai/Mistral-Small-Instruct-2409

Memory needed to run Q4_K_M at 8K tokens

15.8 GB Estimated

File size 13.3 GB Measured

VRAM it fits in (rough guide)

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

See devices and models for 16 GB

The smallest, IQ1_S, needs an estimated 7.31 GB at 8K.

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

Data as of Nov 4, 2026

About Mistral-Small-Instruct-2409

Mistral-Small-Instruct-2409 is a GGUF model based on mistralai/Mistral-Small-Instruct-2409 on Hugging Face. GGUF files are published by MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF. It has about 22.2B parameters, the llama architecture, a maximum context length of 32,768 tokens (values written in the GGUF).

There are 16 quantizations, with file sizes from 4.83 GB (IQ1_S) to 44.5 GB (F16) (actual sizes on Hugging Face).

Running Q4_K_M (13.3 GB) with an 8,192-token context needs an estimated 15.8 GB of memory, and an estimated 21.5 GB at 32,768 tokens. For 8,192 tokens, a GPU with 16 GB or more of VRAM (such as GeForce RTX 5080, GeForce RTX 5070 Ti), or a Mac with 16 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_S (4.83 GB), needs an estimated 7.31 GB at 8,192 tokens.

The original model is licensed under Mistral AI Research License, which does not allow commercial use (limited to non-commercial use such as research or personal use). 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

MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF

Converted and published by a third party About this label

Last checked: Nov 4, 2026

Architecture: llama; Supported before llama.cpp b4409
Parameters: 22.2B
Max context length: 32,768
License: mrl Non-commercial

File size and memory per quantization for Mistral-Small-Instruct-2409 (MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF)
Quantization	File size Measured	Quality guide	4,096	8,192	16,384	32,768 max
IQ1_S	4.83 GB	Very large	6.37 GB	7.31 GB	9.19 GB	12.9 GB
IQ1_M	5.27 GB	Very large	6.81 GB	7.75 GB	9.63 GB	13.4 GB
IQ2_XS	6.65 GB	Very large	8.19 GB	9.13 GB	11.0 GB	14.8 GB
Q2_K	8.27 GB	Very large	9.82 GB	10.8 GB	12.6 GB	16.4 GB
IQ3_XS	9.18 GB	Large	10.7 GB	11.7 GB	13.5 GB	17.3 GB
Q3_K_S	9.64 GB	Large	11.2 GB	12.1 GB	14.0 GB	17.8 GB
Q3_K_M	10.8 GB	Medium	12.3 GB	13.2 GB	15.1 GB	18.9 GB
Q3_K_L	11.7 GB	Medium	13.3 GB	14.2 GB	16.1 GB	19.9 GB
IQ4_XS	11.9 GB	Small	13.5 GB	14.4 GB	16.3 GB	20.1 GB
Q4_K_S	12.7 GB	Medium	14.2 GB	15.1 GB	17.0 GB	20.8 GB
Q4_K_M	13.3 GB	Small	14.9 GB	15.8 GB	17.7 GB	21.5 GB
Q5_K_S	15.3 GB	Small	16.9 GB	17.8 GB	19.7 GB	23.4 GB
Q5_K_M	15.7 GB	Very small	17.3 GB	18.2 GB	20.1 GB	23.8 GB
Q6_K	18.3 GB	Very small	19.8 GB	20.7 GB	22.6 GB	26.4 GB
Q8_0	23.6 GB	Nearly identical	25.2 GB	26.1 GB	28.0 GB	31.8 GB
F16	44.5 GB	Original precision	46.0 GB	47.0 GB	48.9 GB	52.6 GB

Ollama  ollama run hf.co/MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M

llama.cpp  llama-server -hf MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M

Generation speed on common GPUs and Macs (estimated)

A rough guide to how fast Q4_K_M (13.3 GB) generates text. Each token reads the model weights once, so the upper limit is memory bandwidth ÷ file size. The guide is 60–80% of that, where most published measurements fall.

Device	Memory bandwidth	Guide (tokens/s) Estimated	Upper limit
Radeon RX 7600 XT 16 GB	288 GB/s	13–17	22
Radeon RX 9070 16 GB	640 GB/s	29–38	48
GeForce RTX 5090 32 GB	1,792 GB/s	81–107	134
Apple M2 24 GB	100 GB/s	4.5–6.0	7.5
Apple M2 Pro 32 GB	200 GB/s	9.0–12	15
Apple M5 Ultra 96 GB	1,200 GB/s	54–72	90

Estimates only. Prompt processing speed (limited by compute) is not included. Longer contexts and offloading part of the model to the CPU make it slower. Macs are shown only when installed memory is at least 4/3 of what is needed. Devices without published bandwidth (some GeForce cards, etc.) are not shown.

License and commercial use

Mistral AI Research License Non-commercial

Limited to research: personal, scientific or academic research that is non-profit, non-commercial and not directly or indirectly tied to commercial activities. It excludes use in a company's daily work, tests or proofs of concept meant to generate revenue, and distribution or hosting (e.g. SaaS) by commercial entities. Outputs and derivatives carry the same restriction. For commercial use you must ask Mistral AI for a separate license.

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

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/MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M

llama.cpp

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

llama-server -hf MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M

LM Studio

Paste MaziyarPanahi/Mistral-Small-Instruct-2409-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 Mistral-Small-Instruct-2409 (Q4_K_M)?: The file size is 13.3 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 15.8 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can Mistral-Small-Instruct-2409 be used commercially?: The original model's license is Mistral AI Research License (No commercial use).Limited to research: personal, scientific or academic research that is non-profit, non-commercial and not directly or indirectly tied to commercial activities. It excludes use in a company's daily work, tests or proofs of concept meant to generate revenue, and distribution or hosting (e.g. SaaS) by commercial entities. Outputs and derivatives carry the same restriction. For commercial use you must ask Mistral AI for a separate license. Check the original license text before deciding.
Does Mistral-Small-Instruct-2409 run in 8 GB, 12 GB, 16 GB or 24 GB of VRAM?: 8 GB: up to IQ1_M (est. 7.75 GB); 12 GB: up to Q3_K_S (est. 12.1 GB); 16 GB: up to Q4_K_M (est. 15.8 GB); 24 GB: up to Q6_K (est. 20.7 GB) (estimated at 8,192 tokens with an f16 KV cache, files from the most-downloaded source).
What is the command to run Mistral-Small-Instruct-2409 with Ollama?: Run "ollama run hf.co/MaziyarPanahi/Mistral-Small-Instruct-2409-GGUF:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.