Quant	BPW	Head Bits	Size (GB)
2.5_H6	2.5	6	34.98
2.91_H6 (optimized)	2.91	6	40.15
3.0_H6	3.0	6	41.61
3.38_H6 (optimized)	3.38	6	46.33
3.5_H6	3.5	6	48.18
3.85_H6 (optimized)	3.85	6	52.63
4.0_H6	4.0	6	54.80
4.25_H6	4.25	6	58.10
5.0_H6	5.0	6	68.00
6.0_H6	6.0	6	81.20
8.0_H8	8.0	8	107.75

How to Download and Use Quants

You can download quants by targeting specific size using the Hugging Face CLI.

Click for download commands

1. Install huggingface-cli:

pip install -U "huggingface_hub[cli]"

2. Download a specific quant:

huggingface-cli download ArtusDev/Darkhn_GLM-Air-4.5-106B-Animus-V12.0-EXL3 --revision "5.0bpw_H6" --local-dir ./

EXL3 quants can be run with any inference client that supports EXL3, such as TabbyAPI. Refer to documentation for set up instructions.

Quant Requests

See EXL community hub for request guidelines.

Made possible with cloud compute from lium.io