Usage-based — pay for the tokens you use, priced per model.sample data
Billed on input + output tokens at each model's rate. No subscription, no minimums.
Put a monthly USD cap on any API key; live metering stops spend at the limit.
Top up your Neviri AI wallet — a signup credit covers your first requests.
| Model | Provider | Context | Input / 1M | Output / 1M |
|---|---|---|---|---|
| Embed Large | Self-hosted | 8K | $0.02 | — |
| Neviri Open 8B | Self-hosted | 32K | $0.05 | $0.10 |
| Vision Open 12B | Self-hosted | 128K | $0.10 | $0.10 |
| Fast Mini | Resold | 64K | $0.15 | $0.60 |
| Coder 32B | Self-hosted | 128K | $0.18 | $0.18 |
| Neviri Open MoE | Self-hosted | 64K | $0.24 | $0.24 |
| Neviri Open 70B | Self-hosted | 32K | $0.35 | $0.40 |
| Reasoning Mini | Resold | 128K | $0.60 | $2.40 |
| Flagship Chat | Resold | 128K | $2.50 | $10.00 |
| Vision Multimodal | Resold | 128K | $2.50 | $10.00 |
| Reasoning Pro | Resold | 200K | $3.00 | $15.00 |
| Frontier Max | Resold | 200K | $5.00 | $15.00 |
Prices shown are illustrative sample data. Final per-model rates are set in the gateway catalogue at launch; embeddings bill on input only.
