Pricing
Usage-based pricing. Pay for what you use — no platform fees or seat licences.
API Models
Token counting: ~4 characters per token for text. Each time-series data point = 1 token.
| Model | Input | Output | Context | Price |
|---|---|---|---|---|
inertialai-embed-alphaAvailable | Text + Time-Series | Embedding vector 1024 dims | 2048 text tokens 65,534 TS tokens | $0.15 / 1M tokens |
inertialai-forecast-alphaComing Soon | Time-Series + Text (optional) | Time-series forecast target channel | — | Coming Soon |
No output charge for embeddings. No monthly minimum. Token counting: ~4 characters per token for text; each time-series data point = 1 token (max 65,534 per request).
Enterprise
Enterprise Deployment
Private deployment, custom fine-tuning, or high-volume usage agreements.
- Private VPC or on-premise deployment
- Custom fine-tuning on InertialAI Alpha (325M base model)
- SLA & priority support
Frequently Asked Questions
What counts as a token?
For text, approximately 4 characters. For time-series, each data point in the array counts as one token.
When will Forecasting API pricing be available?
Pricing will be published alongside the endpoint launch. Contact us to join the waitlist.
Can I fine-tune the base model?
Yes, as an enterprise option. InertialAI Alpha is available for fine-tuning on proprietary data with private deployment. Contact us to discuss.