How much memory do I need to run GLM-5.3 (UD-IQ1_S)?

The file size is 217 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 218 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.

Does GLM-5.3 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 GLM-5.3 with Ollama?

Run "ollama run hf.co/unsloth/GLM-5.3-GGUF:UD-IQ1_S" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.

zai-org

GLM-5.3

Parameters: 753.9B
Max context length: 1,048,576
Architecture: glm-dsa
License: Conditional

Base model: zai-org/GLM-5.3
License: glm-5.3
llama.cpp: b8036+

Memory needed to run UD-IQ1_S at 8K tokens

218 GB Estimated

File size 217 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 GLM-5.3

GLM-5.3 is a GGUF model based on zai-org/GLM-5.3 on Hugging Face. GGUF files are published by unsloth/GLM-5.3-GGUF. It has about 753.9B parameters, the glm-dsa architecture, a maximum context length of 1,048,576 tokens (values written in the GGUF).

There are 12 quantizations, with file sizes from 217 GB (UD-IQ1_S) to 1,508 GB (BF16) (actual sizes on Hugging Face).

Running UD-IQ1_S (217 GB) with an 8,192-token context needs an estimated 218 GB of memory, and an estimated 221 GB at 32,768 tokens.

The original model is licensed under GLM-5.3 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

unsloth/GLM-5.3-GGUF

Converted and published by a third party About this label

Last checked: Nov 11, 2026

Architecture: glm-dsa; llama.cpp b8036 and later
Parameters: 753.9B
Max context length: 1,048,576
License: glm-5.3 Conditional

File size and memory per quantization for GLM-5.3 (unsloth/GLM-5.3-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-IQ1_S (6 parts)	217 GB	-	218 GB	218 GB	219 GB	221 GB	223 GB	229 GB	241 GB	312 GB
UD-IQ1_M (6 parts)	228 GB	-	230 GB	230 GB	231 GB	232 GB	235 GB	241 GB	253 GB	324 GB
UD-IQ2_M (6 parts)	239 GB	-	240 GB	240 GB	241 GB	242 GB	245 GB	251 GB	263 GB	334 GB
UD-Q2_K_XL (7 parts)	254 GB	-	255 GB	255 GB	256 GB	258 GB	261 GB	267 GB	278 GB	349 GB
UD-IQ3_XXS (7 parts)	282 GB	-	283 GB	283 GB	284 GB	285 GB	288 GB	294 GB	306 GB	377 GB
UD-Q3_K_XL (9 parts)	343 GB	-	344 GB	345 GB	345 GB	347 GB	350 GB	356 GB	367 GB	438 GB
UD-IQ4_XS (9 parts)	365 GB	-	367 GB	367 GB	368 GB	369 GB	372 GB	378 GB	390 GB	460 GB
UD-Q4_K_XL (11 parts)	467 GB	-	469 GB	469 GB	470 GB	471 GB	474 GB	480 GB	492 GB	562 GB
UD-Q5_K_XL (13 parts)	562 GB	-	564 GB	564 GB	565 GB	566 GB	569 GB	575 GB	587 GB	658 GB
UD-Q6_K_XL (16 parts)	684 GB	-	686 GB	686 GB	687 GB	688 GB	691 GB	697 GB	709 GB	779 GB
Q8_0 (17 parts)	801 GB	Nearly identical	803 GB	803 GB	804 GB	805 GB	808 GB	814 GB	826 GB	896 GB
BF16 (33 parts)	1,508 GB	Original precision	1,509 GB	1,510 GB	1,510 GB	1,512 GB	1,515 GB	1,521 GB	1,532 GB	1,603 GB

Ollama  ollama run hf.co/unsloth/GLM-5.3-GGUF:UD-IQ1_S

llama.cpp  llama-server -hf unsloth/GLM-5.3-GGUF:UD-IQ1_S

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)

156.9 154.8–159.2

GLM-5.3 · Values as of Aug 14, 2026

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

License and commercial use

GLM-5.3 License Conditional

Close to MIT: commercial use, modification and redistribution are allowed. Keep the copyright notice and license text, and follow applicable laws. There is one condition aimed at very large operators: 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$10 billion over any 12 months, you must pass Z.AI's security review before commercial use.

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 (glm-dsa) from b8036 (Feb 13, 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-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/unsloth/GLM-5.3-GGUF:UD-IQ1_S

llama.cpp

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

llama-server -hf unsloth/GLM-5.3-GGUF:UD-IQ1_S

LM Studio

Paste unsloth/GLM-5.3-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 GLM-5.3 (UD-IQ1_S)?: The file size is 217 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 218 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
Can GLM-5.3 be used commercially?: The original model's license is GLM-5.3 License (Commercial use with conditions).Close to MIT: commercial use, modification and redistribution are allowed. Keep the copyright notice and license text, and follow applicable laws. There is one condition aimed at very large operators: 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$10 billion over any 12 months, you must pass Z.AI's security review before commercial use. Check the original license text before deciding.
Does GLM-5.3 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 GLM-5.3 with Ollama?: Run "ollama run hf.co/unsloth/GLM-5.3-GGUF:UD-IQ1_S" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.