4-bit Open on Hugging Face
Parameters
3.5B
Weights (est.)
≈ 2.6 GB
Mac memory (est.)
8 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-vlm on an Apple silicon Mac:
pip install mlx-vlmpython -m mlx_vlm.generate --model mlx-community/LocateAnything-3B-4bit --image photo.jpg --prompt "Describe this image"
Details
- Base model
- nvidia/LocateAnything-3B (quantized) · LocateAnything 3B on the radar →
- Vendor
- NVIDIA
- Task
- Vision + text
- Quantisation
- 4-bit
- Licence
- nvidia-license
- Downloads
- 151 in the last 30 days · 5.7K all time · 20 likes
- Published
- 2026-05-29 · updated 2026-05-29 (5mo ago)
- Languages
- en
- Tags
- mlx-vlm, locateanything, vision, object-detection, grounding, nvidia, eagle
Other MLX builds of nvidia/LocateAnything-3B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| LocateAnything-3B-4bit (this one) | 4-bit | ≈ 2.6 GB | 8 GB+ | 151 |
| LocateAnything-3B-8bit | 8-bit | ≈ 4.1 GB | 8 GB+ | 186 |
| LocateAnything-3B-bf16 | bf16 | ≈ 7.0 GB | 16 GB+ | 108 |