skt models
3 skt models on Hugging Face, from 7.3B to 691.7B parameters. At the precision each one is published in, the smallest needs about 16.2 GB of VRAM (A.X-4.0-Light, cheapest live fit: RTX A5000) and the largest about 1160 GB (A.X-K1). The cheapest way to run A.X-4.0-Light is $0.176/hr.
skt models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| A.X-4.0-Light | 7.3B | BF16 | 16.2 GB | RTX A5000 | $0.176/hr |
| A.X-K1 | 519.0B | BF16 | 1160 GB | No live fit | – |
| A.X-K2 | 691.7B | F8_E4M3 | 773 GB | No live fit | – |
VRAM is for the precision the model is published in, with the same overhead and an 8,192-token context assumed on every page; see the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
