4-bit Open on Hugging Face
Parameters
12B
Weights (est.)
≈ 7.0 GB
Mac memory (est.)
16 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/Homunculus-4bit --prompt "Hello"
Details
- Base model
- arcee-ai/Homunculus (quantized)
- Vendor
- arcee-ai
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 20 in the last 30 days · 181 all time · 0 likes
- Published
- 2025-06-03 · updated 2025-06-03 (1y ago)
- Languages
- en
- Tags
- mistral, distillation, /think, /nothink, reasoning-transfer, arcee-ai
Other MLX builds of arcee-ai/Homunculus
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Homunculus-4bit (this one) | 4-bit | ≈ 7.0 GB | 16 GB+ | 20 |
| Homunculus-6bit | 6-bit | ≈ 10.1 GB | 16 GB+ | 22 |
| Homunculus-8bit | 8-bit | ≈ 13.2 GB | 24 GB+ | 20 |
| Homunculus-bf16 | bf16 | ≈ 24.9 GB | 32 GB+ | 27 |