cerebras/GLM-4.7-REAP-218B-A32B needs about 137.4 GB of memory at Q4_K_M (a 133.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 26 quants, from 0.4 GB to 232.2 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 12.3 GB.
GLM-4.7-REAP-218B-A32B – a 218B sparse Mixture‑of‑Experts model
GLM-4.7-REAP-218B-A32B is a 218B parameter model with 32B active experts per token. It uses REAP pruning to reduce memory by 40% while keeping performance close to the full 355B model. The model is licensed under MIT and supports reasoning and tools.
Good at
- Reasoning
- Tools integration
- Sparse expert efficiency
Limits
- No vision support
- Requires large VRAM for full precision
- Only available in GGUF format
Best for
- Local deployment with limited GPU memory
- Academic research on MoE compression
- Code generation and function calling
Who it’s for. Developers and researchers needing a large‑scale language model that fits into constrained hardware. The model is ideal for those who prefer open‑source, MIT‑licensed models and can work with GGUF files in llama.cpp or compatible tools.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.4 GB | 4.0 GB | Fits | Get file → | |
| IQ1_S | 47.9 GB | 51.5 GB | Too big | Get file → | |
| IQ1_M | 50.0 GB (2 parts) | 53.6 GB | Too big | Get file → | |
| IQ2_XXS | 55.4 GB (2 parts) | 59.0 GB | Too big | Get file → | |
| IQ2_XS | 63.1 GB (2 parts) | 66.7 GB | Too big | Get file → | |
| IQ2_S | 63.9 GB (2 parts) | 67.5 GB | Too big | Get file → | |
| IQ2_M | 71.9 GB (2 parts) | 75.5 GB | Too big | Get file → | |
| Q2_K | 79.3 GB (2 parts) | 82.8 GB | Too big | Get file → | |
| Q2_K_L | 80.0 GB (3 parts) | 83.6 GB | Too big | Get file → | |
| IQ3_XXS | 88.3 GB (3 parts) | 91.9 GB | Too big | Get file → | |
| IQ3_XS | 91.8 GB (3 parts) | 95.4 GB | Too big | Get file → | |
| Q3_K_S | 97.1 GB (3 parts) | 100.6 GB | Too big | Get file → | |
| IQ3_M | 101.5 GB (3 parts) | 105.1 GB | Too big | Get file → | |
| Q3_K_M | 101.7 GB (3 parts) | 105.3 GB | Too big | Get file → | |
| Q3_K_L | 104.5 GB (3 parts) | 108.1 GB | Too big | Get file → | |
| Q3_K_XL | 105.2 GB (3 parts) | 108.8 GB | Too big | Get file → | |
| IQ4_XS | 118.2 GB (3 parts) | 121.8 GB | Too big | Get file → | |
| IQ4_NL | 124.8 GB (4 parts) | 128.4 GB | Too big | Get file → | |
| Q4_0 | 126.2 GB (4 parts) | 129.8 GB | Too big | Get file → | |
| Q4_K_S | 129.1 GB (4 parts) | 132.7 GB | Too big | Get file → | |
| Q4_K_M | 133.8 GB (4 parts) | 137.4 GB | Too big | Get file → | |
| Q4_1 | 137.6 GB (4 parts) | 141.2 GB | Too big | Get file → | |
| Q5_K_S | 151.1 GB (4 parts) | 154.7 GB | Too big | Get file → | |
| Q5_K_M | 155.9 GB (4 parts) | 159.5 GB | Too big | Get file → | |
| Q6_K | 179.9 GB (5 parts) | 183.5 GB | Too big | Get file → | |
| Q8_0 | 232.2 GB (6 parts) | 235.8 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_S (4 GB file, 51.5 GB memory at 8 k context). For 16 GB VRAM, use Q4_K_M (133.84 GB file, 137.4 GB memory at 8 k context). For 32 GB VRAM, use Q8_0 (232.18 GB file, 235.8 GB memory at 8 k context). Estimates are approximate.
What GPU do you need for cerebras/GLM-4.7-REAP-218B-A32B?
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: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
- 48 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 4.0 GB in use).
Which cerebras/GLM-4.7-REAP-218B-A32B quant should you choose?
Q4_K_M is the usual starting point for cerebras/GLM-4.7-REAP-218B-A32B: it balances file size and quality well. Here it is a 133.8 GB file.
The Q8_0 file is 232.2 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is the first one in the table at 0.4 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about cerebras/GLM-4.7-REAP-218B-A32B
How much VRAM does cerebras/GLM-4.7-REAP-218B-A32B need?
About 137.4 GB at Q4_K_M with an 8K context window. That is the 133.8 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 cerebras/GLM-4.7-REAP-218B-A32B on an 8 GB GPU?
Yes, with the smaller files: the 0.4 GB file needs about 4.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of cerebras/GLM-4.7-REAP-218B-A32B GGUF?
It depends on the quant: file is 0.4 GB, IQ1_S is 47.9 GB, IQ1_M is 50.0 GB, IQ2_XXS is 55.4 GB, IQ2_XS is 63.1 GB, IQ2_S is 63.9 GB and 20 more.
What context length does cerebras/GLM-4.7-REAP-218B-A32B support?
The GGUF header lists a trained context length of 202,752 tokens (198K). The KV cache grows with the context you use: about 3.1 GB at 8K and 12.3 GB at 32K.
Where can I download cerebras/GLM-4.7-REAP-218B-A32B GGUF?
The files are on Hugging Face in the bartowski/cerebras_GLM-4.7-REAP-218B-A32B-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 cerebras/GLM-4.7-REAP-218B-A32B 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-4.7-REAP-218B-A32B on 8 GB VRAM?
Only the IQ1_S quantization will fit, using about 51.5 GB of memory at 8 k context, which exceeds 8 GB VRAM. The model requires more VRAM for full precision.
Does the model support image input?
No, the model has no vision capabilities.
Is the model free to use?
Yes, it is released under the MIT license.