Pricing
Keep your provider accounts, run serverless inference through Inferock, or choose dedicated capacity. See what you pay in the units you use.
End-to-end encrypted inference, verifiable by design, on our intended confidential routes. See how we qualify a route
Your provider accounts
FREE
Inferock observability on your own provider keys.
Serverless inference
Pay per token
Model rates include provider usage and Inferock service.
10% total service fee on actual supplier cost for every account.
encrypted-inference usage is billed within 30 days
Reserved capacity
By agreement
One capacity quote, Inferock service included.
No seat subscription. Inferock Serve and Inferock Dedicated include observability for their usage. Availability, charging start and service terms are confirmed before activation.
Compare inference options
| How it works | Inferock Watch | Inferock Serve | Inferock Dedicated |
|---|---|---|---|
| Product | Observability | Serverless inference with accountability | Reserved capacity |
| Inference setup | Your provider accounts | Supported models through Inferock | Capacity dedicated to your workload |
| Encryption path | Your provider's connection. The intended confidential route does not apply. | Intended confidential route, designed for end-to-end encryption; qualified per model and configuration. | Intended confidential route, designed for end-to-end encryption; qualified per model and configuration. |
| Price unit | FREE | Total price per model and token type | By agreement |
| Who bills for inference | Your providers | Prepaid through Inferock | Defined in your capacity agreement |
| Observability | Inferock Watch observability is free on your own provider keys; provider usage is billed separately. | Included | Included |
| When charging begins | No Inferock Watch charge on your own provider keys; your providers bill you separately. | As eligible prepaid usage is consumed. | As set in your capacity agreement. Funding covers the capacity commitment before provisioning. |
| Service credits | Separate enrollment required | Eligible failures under agreed terms | Defined by agreement |
Inferock Watch provides observability. A separately enrolled program can provide non-cash credits for future Inferock Serve usage, subject to eligibility, limits and expiry. Inferock Serve provides bounded, non-cash credits for eligible failures under agreed terms. Inferock Dedicated service levels are set in your agreement.
Signup credit, if recorded for your account, is a one-time gift separate from credit earned on eligible Inferock Watch failures. No card, provider connection, paid tier or minimum spend is required to receive it. Credits apply to supplier usage on Inferock Serve, within the available balance and monthly cap. Its service fee stays cash.
Example only
An example of how Inferock Serve rates are shown. It is not an offer, and model and capacity availability through Inferock are not established.
| Product | Input | Cached input | Output |
|---|---|---|---|
| Inferock Watch · observability only | $0 | $0 | $0 |
| Inferock Serve · total rate | $0.33 | $0.066 | $1.32 |
Inferock Watch is free on your own provider keys; your provider bills you separately. Inferock Serve rates include provider usage and Inferock service. Taxes are separate.
H100 80 GB · public-price illustration
$8.00 per GPU-hour supplier-cost illustration
Illustration only, not an Inferock capacity offer. Your quote confirms configuration, commitment, billing unit, total price and service level.
