GLM-5.3-Flash needs about 134.6 GB of memory at its main file (a 84.7 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 2 quants, from 84.7 GB to 124.7 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 197.6 GB.
GLM‑5.3‑Flash RANMA EXL3 GGUF model
GLM‑5.3‑Flash RANMA EXL3 is a 321‑billion‑parameter GLM model repackaged for ranma.cpp. It uses EXL3 quantizations from turboderp and is split into shards. The model is licensed under MIT and requires ranma.cpp `ranma_20261007` or newer.
Good at
- Large context handling
- MoE architecture support
- EXL3 quantization
Limits
- Requires ranma.cpp `ranma_20261007` or newer
- No vision encoder included
- Memory usage high for 8k context
Best for
- Research on large‑context language models
- Testing EXL3 quantization
- Running MoE models in ranma.cpp
Who it’s for. Researchers and developers needing a 321‑billion‑parameter GLM model with EXL3 quantization. The model is suitable for environments that support ranma.cpp and can handle large memory demands.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 84.7 GB (3 parts) | 134.6 GB | Too big | Get file → | |
| — | 124.7 GB (4 parts) | 174.6 GB | Too big | Get file → |
For a 8‑GB VRAM setup, the 84.73‑GB shard set is too large; even the 124.7‑GB set exceeds typical VRAM limits. If host RAM is available, the 84.73‑GB set requires approximately 134.6‑GB memory at 8k context, while the 124.7‑GB set requires about 174.6‑GB. Choose the smaller shard set if memory is constrained.
What GPU do you need for GLM-5.3-Flash?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: no quant fits at an 8K context.
- 32 GB VRAM: no quant fits at an 8K context.
- 48 GB VRAM: no quant fits at an 8K context.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about GLM-5.3-Flash
How much VRAM does GLM-5.3-Flash need?
About 134.6 GB at its main file with an 8K context window. That is the 84.7 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run GLM-5.3-Flash on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 134.6 GB, so you would need a larger GPU or CPU offloading.
What is the file size of GLM-5.3-Flash GGUF?
It depends on the quant: file is 84.7 GB, file is 124.7 GB.
What context length does GLM-5.3-Flash support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 49.4 GB at 8K and 197.6 GB at 32K.
Where can I download GLM-5.3-Flash GGUF?
The files are on Hugging Face in the SaturnHeaven/GLM-5.3-Flash-RANMA-EXL3-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does GLM-5.3-Flash use?
Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run GLM‑5.3‑Flash on 8 GB VRAM?
No. The model requires more than 134.6 GB of memory at 8k context, far exceeding 8 GB VRAM.
Does this model support vision tasks?
No. The vision encoder is a separate mmproj file and is not included in these GGUF shards.
What license does the model use?
The model is licensed under MIT.