# Inferock

> Choose observability with Inferock Watch, serverless inference with accountability through Inferock Serve, or dedicated inference with Inferock Dedicated.

## Key Pages

- [Accountability - sample loss report and clear limits](https://inferock.ai/accountability/): How Inferock separates Inferock Watch visibility from Inferock Serve bounded service credits, with a sanitized sample loss scorecard and clear limits.
- [The Inferock Baseline](https://inferock.ai/baseline/): The requirements a model route must meet to qualify with Inferock, for its exact model, provider route, region and configuration.
- [Careers - Inferock](https://inferock.ai/careers/): Careers at Inferock - we build accountable AI inference infrastructure. Open roles and how to join.
- [Head of Founder Visibility & Partnerships - Inferock Careers](https://inferock.ai/careers/founder-visibility/): Head of Founder Visibility & Partnerships at Inferock - a senior, independent, opportunity-creation role that builds and runs the founder's external presence across media, podcasts, and partnerships.
- [Inferock Dedicated](https://inferock.ai/dedicated/): Run dedicated inference on capacity set aside for your workload, configured and priced by agreement before capacity is provisioned.
- [FAQ - reliable and accountable LLM inference | Inferock](https://inferock.ai/faq/): Answers about reliable and accountable LLM inference, Inferock Watch versus Inferock Serve, eligible failed-call credits, and the receipts that prove what happened.
- [How Inferock works - accountable LLM inference with per-call receipts](https://inferock.ai/how-it-works/): Inferock routes selected model calls through one accountable path, records per-call receipts, and shows whether failures are report-only or eligible for bounded Inferock Serve service credits.
- [Methodology - how Inferock detects and proves failures](https://inferock.ai/methodology/): How Inferock measures failed large language model calls with passive in-path evidence, active probes, and explicit limits for signals that need ground truth or thresholds.
- [Models and routes | Inferock](https://inferock.ai/models/): Explore models routed through Inferock Serve. Compare route details and prices.
- [Pricing - Inferock Watch, Inferock Serve and Inferock Dedicated | Inferock](https://inferock.ai/pricing/): Choose observability with Inferock Watch, serverless inference with accountability through Inferock Serve, or dedicated inference with Inferock Dedicated. See model rates and capacity pricing.
- [Inferock - Privacy](https://inferock.ai/privacy/): Privacy details for Inferock by OpiusAI, including hosted inference data paths, Inferock Watch key handling, per-call evidence storage, and how to contact us.
- [Measure inference through OpenRouter with Inferock](https://inferock.ai/recipes/catalog-losses/): Measure your inference calls locally with inferock-bench. See timing, usage and failure signals, then explore observability with Inferock Watch.
- [Inferock Serve](https://inferock.ai/serve/): Run supported models as serverless inference through Inferock, with independent measurement and service credits for eligible failures.
- [The Inferock Standard - what your AI provider should make good, quantified](https://inferock.ai/standard/): The Inferock Standard defines AI provider SLA accountability, overcharge evidence, time lost, downtime duration, failure taxonomy, evidence grades, and separate money and time ledgers.
- [Inferock - Terms](https://inferock.ai/terms/): Terms for Inferock by OpiusAI, including Inferock Watch and Inferock Serve service modes, customer responsibilities, bounded service credits, and receipt-backed limitations.
- [Inferock Watch](https://inferock.ai/watch/): See how the inference you run on your own provider accounts performs, what it costs and where calls fail.

## Docs

- [Inferock docs](https://inferock.ai/docs/): Inferock documentation index.
- [Account + Inferock Watch | Inferock docs](https://inferock.ai/docs/account-byok/): Set up the account and keys you need to make inference calls through Inferock Watch.
- [Credit promise | Inferock docs](https://inferock.ai/docs/credit-promise/): Understand which remedy applies to your inference option before relying on a credit.
- [First call | Inferock docs](https://inferock.ai/docs/first-call/): Send one inference request and find the corresponding measured call in Inferock Watch.
- [What the gateway forwards | Inferock docs](https://inferock.ai/docs/gateway-forwards/): This page lists exactly what we send, what we change, and what we leave out on purpose.
- [Limits | Inferock docs](https://inferock.ai/docs/limits/): Check the admission limits that apply to your inference requests.
- [Inference pricing | Inferock docs](https://inferock.ai/docs/pricing/): Understand free own-key observability, Inferock Serve service fees and Inferock Dedicated agreement pricing.
- [Quickstart | Inferock docs](https://inferock.ai/docs/quickstart/): Make an inference call with your own provider key, then find the measured call in Inferock Watch.
- [Receipts + ledger | Inferock docs](https://inferock.ai/docs/receipts-ledger/): Platform-fee statements and credit activity have separate records.
- [Set up Inferock Serve | Inferock docs](https://inferock.ai/docs/serve-setup/): Activate Inferock Serve without a provider key, make one inference call, and find it in Calls.
- [Support | Inferock docs](https://inferock.ai/docs/support/): Troubleshoot common setup and gateway errors.

## Answerable Questions

- [How do I get started?](https://inferock.ai/faq/#how-do-i-get-started)
- [What is Inferock?](https://inferock.ai/faq/#what-is-inferock)
- [What is Inferock Dedicated?](https://inferock.ai/faq/#what-is-inferock-dedicated)
- [How are Inferock Watch and Inferock Serve different?](https://inferock.ai/faq/#bring-your-own-keys-managed-different)
- [What happens when an inference call fails?](https://inferock.ai/faq/#what-happens-when-an-inference-call-fails)
- [How does signup credit work?](https://inferock.ai/faq/#how-does-signup-credit-work)
- [What counts as an objective failure?](https://inferock.ai/faq/#objective-failure)
- [What is reported but not automatically credited?](https://inferock.ai/faq/#reported-not-automatically-credited)
- [Is inference through Inferock encrypted end to end?](https://inferock.ai/faq/#is-inference-through-inferock-encrypted-end-to-end)
- [How are my provider keys stored?](https://inferock.ai/faq/#how-are-my-provider-keys-stored)
- [How do I audit my AI or LLM bill?](https://inferock.ai/faq/#audit-ai-llm-bill)
- [Why are my OpenAI or Claude API costs so high?](https://inferock.ai/faq/#why-are-my-openai-or-claude-api-costs-so-high)
- [Can Inferock explain charges when my API usage dashboard shows zero, or when I was not using the API?](https://inferock.ai/faq/#can-inferock-explain-charges-when-my-api-usage-dashboard-shows-zero-or-when-i-was-not-using-the-api)
- [Was I billed for a failed, timed-out, or 504 API call?](https://inferock.ai/faq/#billed-failed-api-call)
- [Can Inferock prove OpenAI overcharged tokens?](https://inferock.ai/faq/#openai-overcharged-tokens)
- [Can Inferock detect an Anthropic billing error?](https://inferock.ai/faq/#anthropic-billing-error)
- [How do I measure Claude or GPT token usage locally?](https://inferock.ai/faq/#measure-claude-gpt-token-usage-locally)
- [Is inferock-bench an LLM cost tracking proxy?](https://inferock.ai/faq/#llm-cost-tracking-proxy)
- [Can inferock-bench stop a runaway agent or retry loop from burning API credits?](https://inferock.ai/faq/#can-inferock-bench-stop-a-runaway-agent-or-retry-loop-from-burning-api-credits)
- [Is this an AI provider reliability benchmark?](https://inferock.ai/faq/#ai-provider-reliability-benchmark)
- [Can Inferock show why my OpenAI or Claude API response is slow?](https://inferock.ai/faq/#can-inferock-show-why-my-openai-or-claude-api-response-is-slow)
- [Can Inferock tell whether a model got worse?](https://inferock.ai/faq/#can-inferock-tell-whether-a-model-got-worse)
- [Does Inferock measure hallucinations?](https://inferock.ai/faq/#does-inferock-measure-hallucinations)
- [What does the loss report include?](https://inferock.ai/faq/#what-does-the-loss-report-include)

## Source

- [GitHub](https://github.com/inferock/inferock-bench)