4-bit Open on Hugging Face
Parameters
37B
Weights (est.)
≈ 21.1 GB
Mac memory (est.)
32 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/K2-Horizon-MoVA-36B-A4B-4bit --prompt "Hello"
Details
- Base model
- IFM/K2-Horizon-MoVA-36B-A4B (quantized)
- Vendor
- IFM
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 586 in the last 30 days · 1.5K all time · 4 likes
- Published
- 2026-09-28 · updated 2026-09-29 (1mo ago)
- Tags
- k2_horizon, quantized, k2-horizon, mova
Other MLX builds of IFM/K2-Horizon-MoVA-36B-A4B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| K2-Horizon-MoVA-36B-A4B-oQ4e | 4-bit | ≈ 21.1 GB | 32 GB+ | 303 |
| K2-Horizon-MoVA-36B-A4B-4bit (this one) | 4-bit | ≈ 21.1 GB | 32 GB+ | 586 |
| K2-Horizon-MoVA-36B-A4B-8bit | 8-bit | ≈ 39.8 GB | 48 GB+ | 256 |
| K2-Horizon-MoVA-36B-A4B-bf16 | bf16 | ≈ 74.9 GB | 96 GB+ | 202 |