maple preview needs about 10.7 GB of memory at its main file (a 9.8 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 3 quants, from 9.8 GB to 12.4 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 1.6 GB.
20B-A1B ternary-weight reasoning LLM optimized for on‑device inference
Maple‑Preview is a 20.2B‑parameter MoE model with 24 layers and 256 experts (8 active). It offers 131,072 token context and is designed for efficient on‑device inference, delivering strong reasoning performance on modest hardware.
Good at
- Reasoning
- Tool use
- On‑device inference
Limits
- No vision support
- Limited agentic training
- Requires 8k context for memory estimates
Best for
- Complex reasoning tasks
- Tool‑augmented workflows
- Low‑latency inference on edge devices
Who it’s for. Developers building reasoning‑heavy applications on local hardware. Researchers interested in MoE architectures and ternary‑weight models. Open‑source enthusiasts looking for a lightweight yet capable LLM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 9.8 GB | 10.7 GB | Too big | Get file → | |
| — | 9.8 GB | 10.7 GB | Too big | Get file → | |
| — | 12.4 GB | 13.3 GB | Too big | Get file → |
The available GGUF files are all "unknown" quantizations. The 9.81 GB files estimate 10.7 GB memory at 8k context, while the 12.44 GB file estimates 13.3 GB memory. For 8 GB VRAM, the 9.81 GB files are the closest fit, but memory usage will still exceed the GPU budget. If higher VRAM is available, the 12.44 GB file offers a larger model size.
What GPU do you need for maple preview?
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: the best quant that fits is the main file (9.8 GB file, about 10.7 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (12.4 GB file, about 13.3 GB in use).
Which maple preview quant should you choose?
If memory is tight, the smallest file is the first one in the table at 9.8 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 maple preview
How much VRAM does maple preview need?
About 10.7 GB at its main file with an 8K context window. That is the 9.8 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 maple preview on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 10.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of maple preview GGUF?
It depends on the quant: file is 9.8 GB, file is 9.8 GB, file is 12.4 GB.
What context length does maple preview support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.4 GB at 8K and 1.6 GB at 32K.
Where can I download maple preview GGUF?
The files are on Hugging Face in the DrMoriarty0/maple-preview repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does maple preview 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 Maple‑Preview on 8 GB VRAM?
The 9.81 GB GGUF files estimate 10.7 GB memory at 8k context, so 8 GB VRAM is insufficient for full inference.
Does Maple‑Preview support vision tasks?
No, the model has no vision capabilities.