How much memory do I need to run Kimi-K3 (UD-Q1_0)?

The file size is 466 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 468 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.

Does Kimi-K3 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 Kimi-K3 with Ollama?

Run "ollama run hf.co/unsloth/Kimi-K3-GGUF:UD-Q1_0" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.

moonshotai

Kimi-K3

Image input

Parameters: 2,779.5B
Max context length: 1,048,576
Architecture: kimi-k3
License: Conditional

Base model: moonshotai/Kimi-K3
License: kimi-k3
llama.cpp: b10448+

Memory needed to run UD-Q1_0 at 8K tokens

468 GB Estimated

File size 466 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

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

Data as of Nov 11, 2026

About Kimi-K3

Kimi-K3 is a GGUF model based on moonshotai/Kimi-K3 on Hugging Face. GGUF files are published by unsloth/Kimi-K3-GGUF. It has about 2,779.5B parameters, the kimi-k3 architecture, a maximum context length of 1,048,576 tokens (values written in the GGUF).

There are 9 quantizations, with file sizes from 466 GB (UD-Q1_0) to 1,561 GB (UD-Q8_K_XL) (actual sizes on Hugging Face).

Running UD-Q1_0 (466 GB) with an 8,192-token context needs an estimated 468 GB of memory, and an estimated 469 GB at 32,768 tokens.

The original model is licensed under Kimi K3 License, which allows commercial use with conditions such as scale or purpose. Check the original license for the terms of use. An mmproj file for image input is also available.

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).

Using image input: This model supports image input. To use images, an mmproj file is loaded as well.

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

unsloth/Kimi-K3-GGUF

Converted and published by a third party About this label

Last checked: Nov 11, 2026

Architecture: kimi-k3; llama.cpp b10448 and later
Parameters: 2,779.5B
Max context length: 1,048,576
License: kimi-k3 Conditional

File size and memory per quantization for Kimi-K3 (unsloth/Kimi-K3-GGUF)
Quantization	File size Measured	Quality guide	4,096	8,192	16,384	32,768	65,536	131,072	262,144	1,048,576 max
UD-Q1_0 (11 parts)	466 GB	-	468 GB	468 GB	468 GB	469 GB	470 GB	471 GB	475 GB	497 GB
UD-TQ1_0 (12 parts)	509 GB	-	510 GB	510 GB	511 GB	511 GB	512 GB	514 GB	517 GB	539 GB
UD-TQ2_0 (13 parts)	551 GB	-	553 GB	553 GB	553 GB	554 GB	555 GB	556 GB	560 GB	582 GB
UD-IQ1_S (14 parts)	594 GB	-	595 GB	596 GB	596 GB	596 GB	597 GB	599 GB	603 GB	624 GB
UD-IQ1_M (15 parts)	649 GB	-	650 GB	650 GB	651 GB	651 GB	652 GB	654 GB	657 GB	679 GB
UD-IQ2_XXS (16 parts)	711 GB	-	713 GB	713 GB	713 GB	713 GB	714 GB	716 GB	720 GB	741 GB
UD-Q2_K_XL (19 parts)	861 GB	-	863 GB	863 GB	863 GB	864 GB	864 GB	866 GB	870 GB	892 GB
UD-Q4_K_XL (32 parts)	1,509 GB	-	1,510 GB	1,510 GB	1,510 GB	1,511 GB	1,512 GB	1,514 GB	1,517 GB	1,539 GB
UD-Q8_K_XL (34 parts)	1,561 GB	-	1,563 GB	1,563 GB	1,563 GB	1,563 GB	1,564 GB	1,566 GB	1,570 GB	1,591 GB

The mmproj file for image input (F16) is 0.90 GB Measured . It is not included in the memory above, so add it if you use images.

Ollama  ollama run hf.co/unsloth/Kimi-K3-GGUF:UD-Q1_0

llama.cpp  llama-server -hf unsloth/Kimi-K3-GGUF:UD-Q1_0

Third-party evaluations

Values from other organizations for the original model before quantization. Not this site's own rating. Values are for the original model before quantization. GGUF quantization (especially small quantizations) can lower performance.

Capabilities (Epoch AI)

158.5 156.5–160.7

Kimi K3 · Values as of Jul 16, 2026

Source: Epoch Capabilities Index (Epoch AI) (CC BY 4.0). Original leaderboard · See in ranking

License and commercial use

Kimi K3 License Conditional

Commercial use, modification and redistribution are allowed. If you run a "Model as a Service" business (giving third parties inference or fine-tuning access) and you and your affiliates earn more than US$20 million over any 12 months, you must sign a separate agreement with Moonshot AI before commercial use. Products or services with more than 100 million monthly active users or more than US$20 million monthly revenue must prominently display "Kimi K3". Internal use, and use through Moonshot AI's official products or certified inference partners, is exempt from these two conditions.

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 (kimi-k3) from b10448 (Aug 15, 2026) 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 UD-Q1_0. 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/unsloth/Kimi-K3-GGUF:UD-Q1_0

llama.cpp

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

llama-server -hf unsloth/Kimi-K3-GGUF:UD-Q1_0

LM Studio

Paste unsloth/Kimi-K3-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 Kimi-K3 (UD-Q1_0)?: The file size is 466 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 468 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can Kimi-K3 be used commercially?: The original model's license is Kimi K3 License (Commercial use with conditions).Commercial use, modification and redistribution are allowed. If you run a "Model as a Service" business (giving third parties inference or fine-tuning access) and you and your affiliates earn more than US$20 million over any 12 months, you must sign a separate agreement with Moonshot AI before commercial use. Products or services with more than 100 million monthly active users or more than US$20 million monthly revenue must prominently display "Kimi K3". Internal use, and use through Moonshot AI's official products or certified inference partners, is exempt from these two conditions. Check the original license text before deciding.
Does Kimi-K3 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 Kimi-K3 with Ollama?: Run "ollama run hf.co/unsloth/Kimi-K3-GGUF:UD-Q1_0" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.