AesSedai/GLM-5.3-GGUF overview
This repo contains specialized MoE quants for zai org/GLM 5.3 BF16. The idea being that given the huge size of the FFN tensors compared to the rest of the tens…
Runs locally from ~9.0 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| IQ2_S/GLM-5.3-IQ2_S-00001-of-00007.gguf | GGUF | IQ2_S | 9.0 MB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00002-of-00007.gguf | GGUF | IQ2_S | 46.48 GB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00003-of-00007.gguf | GGUF | IQ2_S | 45.82 GB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00004-of-00007.gguf | GGUF | IQ2_S | 45.82 GB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00005-of-00007.gguf | GGUF | IQ2_S | 45.78 GB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00006-of-00007.gguf | GGUF | IQ2_S | 45.82 GB | Download |
| IQ2_S/GLM-5.3-IQ2_S-00007-of-00007.gguf | GGUF | IQ2_S | 11.61 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00001-of-00007.gguf | GGUF | IQ3_S | 9.0 MB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00002-of-00007.gguf | GGUF | IQ3_S | 46.54 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00003-of-00007.gguf | GGUF | IQ3_S | 46.23 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00004-of-00007.gguf | GGUF | IQ3_S | 46.39 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00005-of-00007.gguf | GGUF | IQ3_S | 46.03 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00006-of-00007.gguf | GGUF | IQ3_S | 46.23 GB | Download |
| IQ3_S/GLM-5.3-IQ3_S-00007-of-00007.gguf | GGUF | IQ3_S | 34.51 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00001-of-00009.gguf | GGUF | IQ4_XS | 9.0 MB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00002-of-00009.gguf | GGUF | IQ4_XS | 45.70 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00003-of-00009.gguf | GGUF | IQ4_XS | 45.35 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00004-of-00009.gguf | GGUF | IQ4_XS | 45.46 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00005-of-00009.gguf | GGUF | IQ4_XS | 46.51 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00006-of-00009.gguf | GGUF | IQ4_XS | 45.61 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00007-of-00009.gguf | GGUF | IQ4_XS | 46.51 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00008-of-00009.gguf | GGUF | IQ4_XS | 45.65 GB | Download |
| IQ4_XS/GLM-5.3-IQ4_XS-00009-of-00009.gguf | GGUF | IQ4_XS | 21.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00001-of-00011.gguf | GGUF | Q4_K_M | 9.0 MB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00002-of-00011.gguf | GGUF | Q4_K_M | 44.93 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00003-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00004-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00005-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00006-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00007-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00008-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00009-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00010-of-00011.gguf | GGUF | Q4_K_M | 45.22 GB | Download |
| Q4_K_M/GLM-5.3-Q4_K_M-00011-of-00011.gguf | GGUF | Q4_K_M | 30.23 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00001-of-00013.gguf | GGUF | Q5_K_M | 9.0 MB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00002-of-00013.gguf | GGUF | Q5_K_M | 46.56 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00003-of-00013.gguf | GGUF | Q5_K_M | 45.16 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00004-of-00013.gguf | GGUF | Q5_K_M | 45.34 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00005-of-00013.gguf | GGUF | Q5_K_M | 45.54 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00006-of-00013.gguf | GGUF | Q5_K_M | 45.14 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00007-of-00013.gguf | GGUF | Q5_K_M | 45.34 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00008-of-00013.gguf | GGUF | Q5_K_M | 45.54 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00009-of-00013.gguf | GGUF | Q5_K_M | 45.14 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00010-of-00013.gguf | GGUF | Q5_K_M | 45.34 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00011-of-00013.gguf | GGUF | Q5_K_M | 45.54 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00012-of-00013.gguf | GGUF | Q5_K_M | 45.14 GB | Download |
| Q5_K_M/GLM-5.3-Q5_K_M-00013-of-00013.gguf | GGUF | Q5_K_M | 23.27 GB | Download |
| imatrix.gguf | GGUF | GGUF | 1.05 GB | Download |
Model Details
Model README
---
base_model:
- zai-org/GLM-5.3-BF16
---
This repo contains specialized MoE-quants for zai-org/GLM-5.3-BF16. The idea being that given the huge size of the FFN tensors compared to the rest of the tensors in the model, it should be possible to achieve a better quality while keeping the overall size of the entire model smaller compared to a similar naive quantization. To that end, the quantization type default is kept in high quality and the FFN UP + FFN GATE tensors are quanted down along with the FFN DOWN tensors.
| Quant | Size | Mixture | PPL | 1-(Mean PPL(Q)/PPL(base)) | KLD |
| :----- | :-------------------- | :------------------------------- | :------------------ | :------------------------ | :------------------ |
| Q8_0 | 745.77 GiB (8.50 BPW) | Q8_0 | 2.674779 ± 0.013836 | +0.0174% | 0.013080 ± 0.000123 |
| Q5_K_M | 523.07 GiB (5.96 BPW) | Q8_0 / Q5_K / Q5_K / Q6_K | 2.680808 ± 0.013885 | +0.2428% | 0.020301 ± 0.000174 |
| Q4_K_M | 436.94 GiB (4.98 BPW) | Q8_0 / Q4_K / Q4_K / Q5_K | 2.699454 ± 0.013914 | +0.9400% | 0.037094 ± 0.000294 |
| IQ4_XS | 342.01 GiB (3.90 BPW) | Q8_0 / IQ3_S / IQ3_S / IQ4_XS | 2.841042 ± 0.014851 | +6.2344% | 0.101583 ± 0.000741 |
| IQ3_S | 265.93 GiB (3.03 BPW) | Q6_K / IQ2_S / IQ2_S / IQ3_S | 3.257199 ± 0.017856 | +21.7956% | 0.275247 ± 0.001717 |
| IQ2_S | 241.32 GiB (2.75 BPW) | Q6_K / IQ2_XS / IQ2_XS / IQ3_XXS | 3.512105 ± 0.019532 | +31.3273% | 0.374183 ± 0.002137 |
Run AesSedai/GLM-5.3-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models