PAW 24B X3.1 needs about 11.5 GB of memory at its main file (a 8.3 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 4 quants, from 6.9 GB to 11.2 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 10.7 GB.
PAW 24B X3.1 – a 4.63B‑parameter MoE reasoning model
PAW 24B X3.1 is a 4.63‑billion‑parameter LFM2MOE model built on LiquidAI/LFM2.5‑24B‑A2B. It supports 131,072‑token context and is designed for on‑device reasoning with function calling. The model is available in four GGUF files ranging from 6.9 GB to 11.2 GB.
Good at
- Large context reasoning
- Function calling
- On‑device inference
Limits
- No vision support
- Requires 24 GB RTX 3090 or smaller for full model
- License is other
Best for
- Agentic workflows
- Tool‑calling tasks
- Long‑form reasoning
Who it’s for. Developers building private agents that need deep reasoning and tool use on local hardware. The model is suitable for users who can host it on a GPU with at least 24 GB VRAM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 6.9 GB | 10.1 GB | Too big | Get file → | |
| — | 8.3 GB | 11.5 GB | Too big | Get file → | |
| — | 9.8 GB | 13.0 GB | Too big | Get file → | |
| — | 11.2 GB | 14.4 GB | Too big | Get file → |
For 8 GB VRAM use the 6.9 GB file, which estimates 10.1 GB memory at 8 k context. For 12 GB VRAM use the 8.33 GB file, estimating 11.5 GB memory. For 16 GB VRAM use the 9.77 GB file, estimating 13 GB memory. For 24 GB VRAM use the 11.2 GB file, estimating 14.4 GB memory. Estimates are approximate.
What GPU do you need for PAW 24B X3.1?
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 (8.3 GB file, about 11.5 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (11.2 GB file, about 14.4 GB in use).
Which PAW 24B X3.1 quant should you choose?
If memory is tight, the smallest file is the first one in the table at 6.9 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 PAW 24B X3.1
How much VRAM does PAW 24B X3.1 need?
About 11.5 GB at its main file with an 8K context window. That is the 8.3 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 PAW 24B X3.1 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 10.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of PAW 24B X3.1 GGUF?
It depends on the quant: file is 6.9 GB, file is 8.3 GB, file is 9.8 GB, file is 11.2 GB.
What context length does PAW 24B X3.1 support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 2.7 GB at 8K and 10.7 GB at 32K.
Where can I download PAW 24B X3.1 GGUF?
The files are on Hugging Face in the lackonendes/PAW-24B-X3.1 repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does PAW 24B X3.1 use?
Hugging Face lists the license as other. 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 PAW 24B X3.1 on a 24 GB RTX 3090?
Yes, the 11.2 GB GGUF file runs on a 24 GB RTX 3090 or smaller. The model uses 14.4 GB memory at 8 k context.
Does PAW 24B X3.1 support vision?
No, the model is text‑only and does not support vision.
What is the context length of PAW 24B X3.1?
The model supports a context length of 131,072 tokens.