Skip to main content
The MetrixLLM dashboard is your central control panel for managing your AI gateway. Everything you configure here is applied automatically to live traffic — no deployments required. The sidebar organizes your workspace into the following sections:

Request Logs

Inspect every individual LLM call with full payloads, costs, and latency.

Cost Analytics

Monitor and attribute AI spend across models, providers, and API keys.

Prompts

Create, version, and deploy prompts without touching your code.

Playground

Interactively test prompts and compare models in a live IDE.

Guardrails

Configure content and safety rules applied at the gateway layer.

Routing Rules

Define fallback chains and retry logic for your requests.

API Keys

Create and revoke keys for your applications.

Settings

Configure workspace-level options like caching and data retention.

​
Workspaces

At the top of the sidebar you will see a Workspace selector. Each workspace is a fully isolated environment with its own API keys, logs, routing rules, and provider credentials. A common setup is to have one workspace per environment:
  • Production
  • Staging
  • Development
Switching workspaces updates all pages to show only that workspace’s data.