Training, sampling, and inference on River Cloud are billed per token. Prices are in USD per 1M tokens.
Training & sampling
| Model | Context | Prompt | Cached | Completion | Training |
|---|---|---|---|---|---|
| Loading prices… | |||||
Prompt, cached, and completion are sampling prices · Checkpoint storage billed at $0.10 / GB / month
Inference
Call open models through an OpenAI-compatible API with your River API key. See the inference quickstart.
| Model | Prompt | Cached | Completion |
|---|---|---|---|
| Loading prices… | |||
Custom and on-prem
Longer context lengths, additional models, and deployment on your own GPUs are priced on request. Talk to our team.
Preview rates, subject to change.