Inferock

Inferock is an LLM inference service that attaches per-call evidence to model requests. You route calls through it on your own provider keys, on Inferock Serve or on Inferock Dedicated. It checks responses against your app's requirements and shows quality findings, timing, usage and available cost data. A recorded failure can be traced to the call behind it, and a monthly loss report groups failed calls by failure type, event count, estimated cost, evidence and remedy status.

There are three ways to run. Inferock Watch observes calls made through your own provider accounts. Inferock Serve is managed serverless inference that you prepay, with no provider API key required. Inferock Dedicated is reserved capacity with configuration, price and service level agreed before provisioning. One endpoint covers multiple providers; measured public benchmark support today is OpenAI, Anthropic, Gemini Developer API and pinned OpenRouter endpoints, and other providers are described as a design capability.

Inferock Watch is free on your own provider keys, and your providers bill you separately. Inferock Serve is priced per token with a 10% total service fee on actual supplier cost for every account, and it includes non-cash service credits for eligible failures under agreed terms. Inferock Dedicated is priced by agreement through one capacity quote. The pricing page states there is no seat subscription, and no card or minimum spend is required to receive a signup credit.

Before buying, note that the rates shown are examples, not an offer, and availability, charging start and service terms are confirmed before activation. The confidential route is described as intended: each route still requires qualification evidence for its exact model and configuration, and the protection does not apply to ordinary Inferock Watch connections. Service credits are non-cash and bounded. Support is by email and the page says replies come within one business day.