How much memory do I need to run DeepSeek-R1-0528 (Q4_K_M)?

The file size is 409 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 411 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.

Does DeepSeek-R1-0528 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 DeepSeek-R1-0528 with Ollama?

Run "ollama run hf.co/bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.

deepseek-ai

DeepSeek-R1-0528

Parameters: 671B
Max context length: 163,840
Architecture: deepseek2
License: Commercial OK

Base model: deepseek-ai/DeepSeek-R1-0528
License: mit

Memory needed to run Q4_K_M at 8K tokens

411 GB Estimated

File size 409 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

The smallest, IQ1_M, needs an estimated 149 GB at 8K.

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

Data as of Oct 13, 2026

About DeepSeek-R1-0528

DeepSeek-R1-0528 is a GGUF model based on deepseek-ai/DeepSeek-R1-0528 on Hugging Face. GGUF files are published by bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF. It has about 671B parameters, the deepseek2 architecture, a maximum context length of 163,840 tokens (values written in the GGUF).

There are 24 quantizations, with file sizes from 147 GB (IQ1_M) to 713 GB (Q8_0) (actual sizes on Hugging Face).

Running Q4_K_M (409 GB) with an 8,192-token context needs an estimated 411 GB of memory, and an estimated 412 GB at 32,768 tokens. The smallest, IQ1_M (147 GB), needs an estimated 149 GB at 8,192 tokens.

The original model is licensed under MIT License, 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

bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF

Converted and published by a third party About this label

Last checked: Oct 13, 2026

Architecture: deepseek2; Supported before llama.cpp b4409
Parameters: 671.0B
Max context length: 163,840
License: mit Commercial OK

File size and memory per quantization for DeepSeek-R1-0528 (bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF)
Quantization	File size Measured	Quality guide	4,096	8,192	16,384	32,768	65,536	131,072	163,840 max
IQ1_M (4 parts)	147 GB	Very large	149 GB	149 GB	149 GB	151 GB	153 GB	157 GB	160 GB
IQ2_XXS (5 parts)	164 GB	Very large	165 GB	165 GB	166 GB	167 GB	169 GB	174 GB	176 GB
IQ2_XS (5 parts)	188 GB	Very large	189 GB	190 GB	190 GB	192 GB	194 GB	198 GB	201 GB
IQ2_S (5 parts)	190 GB	Very large	191 GB	191 GB	192 GB	193 GB	195 GB	200 GB	202 GB
IQ2_M (6 parts)	215 GB	Large	216 GB	216 GB	217 GB	218 GB	220 GB	225 GB	227 GB
Q2_K (7 parts)	238 GB	Very large	239 GB	239 GB	240 GB	241 GB	243 GB	248 GB	250 GB
Q2_K_L (7 parts)	239 GB	-	240 GB	240 GB	241 GB	242 GB	244 GB	249 GB	251 GB
IQ3_XXS (7 parts)	268 GB	Large	269 GB	269 GB	270 GB	271 GB	273 GB	278 GB	280 GB
IQ3_XS (8 parts)	277 GB	Large	278 GB	279 GB	279 GB	280 GB	283 GB	287 GB	289 GB
Q3_K_S (8 parts)	293 GB	Large	294 GB	295 GB	295 GB	296 GB	299 GB	303 GB	306 GB
IQ3_M (8 parts)	308 GB	Medium	309 GB	309 GB	310 GB	311 GB	313 GB	318 GB	320 GB
Q3_K_M (8 parts)	308 GB	Medium	309 GB	309 GB	310 GB	311 GB	313 GB	318 GB	320 GB
Q3_K_L (9 parts)	320 GB	Medium	321 GB	321 GB	322 GB	323 GB	325 GB	330 GB	332 GB
Q3_K_XL (9 parts)	321 GB	-	322 GB	322 GB	322 GB	324 GB	326 GB	331 GB	333 GB
IQ4_XS (10 parts)	360 GB	Small	361 GB	361 GB	362 GB	363 GB	365 GB	370 GB	372 GB
IQ4_NL (10 parts)	380 GB	Small	382 GB	382 GB	382 GB	384 GB	386 GB	390 GB	393 GB
Q4_0 (10 parts)	386 GB	Medium	388 GB	388 GB	388 GB	390 GB	392 GB	396 GB	399 GB
Q4_K_S (11 parts)	394 GB	Medium	395 GB	396 GB	396 GB	397 GB	400 GB	404 GB	406 GB
Q4_K_M (11 parts)	409 GB	Small	410 GB	411 GB	411 GB	412 GB	415 GB	419 GB	422 GB
Q4_1 (11 parts)	421 GB	Medium	422 GB	422 GB	423 GB	424 GB	426 GB	431 GB	433 GB
Q5_K_S (12 parts)	463 GB	Small	464 GB	464 GB	465 GB	466 GB	468 GB	473 GB	475 GB
Q5_K_M (13 parts)	478 GB	Very small	479 GB	480 GB	480 GB	481 GB	484 GB	488 GB	491 GB
Q6_K (15 parts)	552 GB	Very small	554 GB	554 GB	554 GB	556 GB	558 GB	562 GB	565 GB
Q8_0 (20 parts)	713 GB	Nearly identical	714 GB	715 GB	715 GB	716 GB	719 GB	723 GB	726 GB

Ollama  ollama run hf.co/bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M

llama.cpp  llama-server -hf bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M

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.

Human preference (LMArena)

1,421 1,416–1,427

DeepSeek-R1 (0528) · 18,153 votes · Values as of Nov 9, 2026

Source: LMArena (CC BY 4.0). Original leaderboard · See in ranking

Hard science (GPQA)

76.3% 71.6%–81.0%

DeepSeek-R1-0528 · Values as of May 29, 2025

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

License and commercial use

MIT License Commercial OK

Allows commercial use, modification and redistribution. When redistributing, keep the copyright notice and license text.

Based on the license on the original model's Hugging Face page. 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 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/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M

llama.cpp

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

llama-server -hf bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M

LM Studio

Paste bartowski/deepseek-ai_DeepSeek-R1-0528-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 DeepSeek-R1-0528 (Q4_K_M)?: The file size is 409 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 411 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can DeepSeek-R1-0528 be used commercially?: The original model's license is MIT License (Commercial use allowed).Allows commercial use, modification and redistribution. When redistributing, keep the copyright notice and license text. Check the original license text before deciding.
Does DeepSeek-R1-0528 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 DeepSeek-R1-0528 with Ollama?: Run "ollama run hf.co/bartowski/deepseek-ai_DeepSeek-R1-0528-GGUF:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.