ArliAI GLM 4.5 Air Derestricted REAP50 Creative needs about 62.7 GB of memory at Q8_0 (a 60.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. Longer context needs more: at 32K tokens the KV cache alone adds about 6.2 GB.
ArliAI GLM 4.5 Air Derestricted REAP50 Creative – 57B parameter model
ArliAI GLM 4.5 Air Derestricted REAP50 Creative is a 57B parameter GLM model with a 131072 token context window. It is a 50% REAP of the derestricted GLM-4.5-Air, optimized for creative English tasks in memory‑constrained environments.
Good at
- Creative writing
- English language tasks
- Reasoning with large context
Limits
- Requires large VRAM – 60.62 GB file, 62.7 GB memory at 8k context
- No vision support
- Unverified loading and quantization
Best for
- Short story generation
- Poetry creation
- Creative brainstorming
Who it’s for. Developers and researchers needing a large‑context, creative language model on consumer hardware. The model is suitable for applications where refusal behavior is undesirable and memory constraints are tight.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 60.6 GB (4 parts) | 62.7 GB | Too big | Get file → |
The only available GGUF quant is Q8_0, 60.62 GB. It requires approximately 62.7 GB of VRAM at 8k context. If your GPU has 64 GB or more, this quant can be loaded. No smaller quant is currently available.
What GPU do you need for ArliAI GLM 4.5 Air Derestricted REAP50 Creative?
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.
Which ArliAI GLM 4.5 Air Derestricted REAP50 Creative quant should you choose?
Q8_0 (60.6 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about ArliAI GLM 4.5 Air Derestricted REAP50 Creative
How much VRAM does ArliAI GLM 4.5 Air Derestricted REAP50 Creative need?
About 62.7 GB at Q8_0 with an 8K context window. That is the 60.6 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 ArliAI GLM 4.5 Air Derestricted REAP50 Creative on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 62.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of ArliAI GLM 4.5 Air Derestricted REAP50 Creative GGUF?
It depends on the quant: Q8_0 is 60.6 GB.
What context length does ArliAI GLM 4.5 Air Derestricted REAP50 Creative support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.5 GB at 8K and 6.2 GB at 32K.
Where can I download ArliAI GLM 4.5 Air Derestricted REAP50 Creative GGUF?
The files are on Hugging Face in the Koshkasa/ArliAI_GLM-4.5-Air-Derestricted_REAP50_Creative repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does ArliAI GLM 4.5 Air Derestricted REAP50 Creative 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 this model on 8 GB VRAM?
No. The Q8_0 quant requires about 62.7 GB of VRAM at 8k context, far exceeding 8 GB.
Does this model support vision tasks?
No. The model is purely text‑based and has no vision capabilities.
Is the model fully tested?
No. The release notes state the model is currently untested and may produce gibberish or fail to load.