Laguna models
4 Laguna models on Hugging Face, from 33.4B to 117.6B parameters. At the precision each one is published in, the smallest needs about 74.8 GB of VRAM (Laguna-XS-2.1, cheapest live fit: A100) and the largest about 263 GB (Laguna-S-2.1, cheapest live fit: RTX A6000). The cheapest way to run Laguna-XS-2.1 is $1.21/hr.
Laguna models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Laguna-XS-2.1 | 33.4B | BF16 | 74.8 GB | A100 | $1.21/hr |
| Laguna-XS.2 | 33.4B | BF16 | 74.8 GB | A100 | $1.21/hr |
| Laguna-S-2.1 | 117.6B | BF16 | 263 GB | RTX A6000 × 6 | $2.18/hr |
| Laguna-S-2.1-FP8 | 117.6B | F8_E4M3 | 131 GB | RTX 4090 × 3 | $1.32/hr |
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.
