Homunculus needs about 9.6 GB of memory at Q4_K_M (a 7.8 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 31 quants, from 3.4 GB (TQ1_0) to 24.9 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 5.4 GB.
Homunculus is a 12.5B-parameter Qwen model in GGUF format, quantized by Mungert. Its trained context length is 128K tokens. The repository has 31 quant files, from 3.4 GB (TQ1_0) to 24.9 GB (BF16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| TQ1_0 | 3.4 GB | 5.2 GB | Fits | Get file → | |
| IQ1_S | 3.9 GB | 5.7 GB | Fits | Get file → | |
| TQ2_0 | 3.9 GB | 5.7 GB | Fits | Get file → | |
| IQ1_M | 4.1 GB | 6.0 GB | Fits | Get file → | |
| IQ2_XXS | 4.2 GB | 6.1 GB | Fits | Get file → | |
| IQ2_XS | 4.6 GB | 6.4 GB | Fits | Get file → | |
| IQ2_S | 4.7 GB | 6.5 GB | Fits | Get file → | |
| Q2_K_S | 4.8 GB | 6.6 GB | Fits | Get file → | |
| IQ2_M | 4.9 GB | 6.7 GB | Fits | Get file → | |
| Q2_K_M | 5.2 GB | 7.1 GB | Fits | Get file → | |
| IQ3_XXS | 5.3 GB | 7.1 GB | Fits | Get file → | |
| IQ3_XS | 5.5 GB | 7.4 GB | Tight | Get file → | |
| IQ3_S | 5.8 GB | 7.6 GB | Tight | Get file → | |
| Q3_K_S | 5.8 GB | 7.7 GB | Tight | Get file → | |
| IQ3_M | 5.9 GB | 7.7 GB | Tight | Get file → | |
| Q3_K_M | 6.5 GB | 8.3 GB | Too big | Get file → | |
| IQ4_XS | 6.9 GB | 8.7 GB | Too big | Get file → | |
| Q4_0 | 7.0 GB | 8.9 GB | Too big | Get file → | |
| IQ4_NL | 7.2 GB | 9.1 GB | Too big | Get file → | |
| Q4_K_S | 7.5 GB | 9.4 GB | Too big | Get file → | |
| Q4_K_M | 7.8 GB | 9.6 GB | Too big | Get file → | |
| Q4_1 | 7.8 GB | 9.6 GB | Too big | Get file → | |
| Q5_0 | 8.6 GB | 10.4 GB | Too big | Get file → | |
| Q5_K_S | 8.9 GB | 10.7 GB | Too big | Get file → | |
| Q5_K_M | 9.0 GB | 10.9 GB | Too big | Get file → | |
| Q5_1 | 9.4 GB | 11.2 GB | Too big | Get file → | |
| Q6_K_M | 10.2 GB | 12.1 GB | Too big | Get file → | |
| Q8_0 | 13.2 GB | 15.1 GB | Too big | Get file → | |
| BF16 | 18.4 GB | 20.3 GB | Too big | Get file → | |
| F16 | 18.4 GB | 20.3 GB | Too big | Get file → | |
| BF16 | 24.9 GB | 26.8 GB | Too big | Get file → |
What GPU do you need for Homunculus?
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: the best quant that fits is IQ1_M (4.1 GB file, about 6.0 GB in use).
- 8 GB VRAM: the best quant that fits is IQ3_M (5.9 GB file, about 7.7 GB in use).
- 12 GB VRAM: the best quant that fits is Q5_1 (9.4 GB file, about 11.2 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (13.2 GB file, about 15.1 GB in use).
- 24 GB VRAM: the best quant that fits is BF16 (18.4 GB file, about 20.3 GB in use).
- 32 GB VRAM: the best quant that fits is BF16 (24.9 GB file, about 26.8 GB in use).
Which Homunculus quant should you choose?
Q4_K_M is the usual starting point for Homunculus: it balances file size and quality well. Here it is a 7.8 GB file.
The Q8_0 file is 13.2 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 TQ1_0 at 3.4 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Homunculus
How much VRAM does Homunculus need?
About 9.6 GB at Q4_K_M with an 8K context window. That is the 7.8 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 Homunculus on an 8 GB GPU?
Yes, with IQ3_M: the 5.9 GB file needs about 7.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Homunculus GGUF?
It depends on the quant: TQ1_0 is 3.4 GB, IQ1_S is 3.9 GB, TQ2_0 is 3.9 GB, IQ1_M is 4.1 GB, IQ2_XXS is 4.2 GB, IQ2_XS is 4.6 GB and 25 more.
What context length does Homunculus support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.3 GB at 8K and 5.4 GB at 32K.
Where can I download Homunculus GGUF?
The files are on Hugging Face in the Mungert/Homunculus-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 Homunculus use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.