Models
Open models, one key.
10 open models, read live from the upstream router. Each is metered at the list price its paid hosts charge; the second price is what a request costs with DOSE bought at the deepest discount in the book right now, fee included.
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning | ||
| Inkling Smallthinkingmachines/inkling-small | ||
| Laguna S 2.1poolside/laguna-s-2.1 | ||
| Inklingthinkingmachines/inkling | ||
| Laguna XS 2.1poolside/laguna-xs-2.1 | ||
| Nemotron 3.5 Content Safetynvidia/nemotron-3.5-content-safety | ||
| Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b | ||
| Gemma 4 26B A4B google/gemma-4-26b-a4b-it | ||
| Gemma 4 31Bgoogle/gemma-4-31b-it | ||
| Nemotron 3 Supernvidia/nemotron-3-super-120b-a12b |
Veleno does not host these models. It relays requests to the upstream router's free capacity for each one, so the upstream's limits apply; a request the upstream refuses is not charged.