Metrics & Analytics

Routero’s built-in analytics give every stakeholder the view they need — developers see their key’s usage, teams see their spend, admins see the full workspace.

Dashboard

The Routero admin dashboard (platform.routero.ai) provides:

  • Usage overview — total requests, tokens, cost, and latency trends (1h / 24h / 7d / 30d)
  • Model breakdown — requests and cost by model and provider
  • Team breakdown — per-team spend with budget remaining indicators
  • Key activity — per-key request counts and cost
  • Routing decisions — provider distribution, fallback frequency, error rates
  • Budget alerts — active spend warnings and blocks

Analytics API

Pull usage data programmatically:

# Org-level daily spend
GET /billing/daily-spend?start_date=2026-06-01&end_date=2026-06-30

# Team activity
GET /team/daily/activity?team_id=data-science&start_date=2026-06-01

# User activity
GET /user/daily/activity?user_id=alice@company.com

# Per-key spend
GET /key/info?key_hash=sk-...

Prometheus metrics

When Prometheus is enabled (in Docker Compose or as an ECS task), Routero exposes a /metrics endpoint with:

  • litellm_request_total — request count by model, provider, status
  • litellm_request_duration_seconds — latency histograms
  • litellm_tokens_total — input/output token counts
  • litellm_cost_total — spend in USD

Scrape config:

scrape_configs:
  - job_name: routero
    static_configs:
      - targets: ["routero-proxy:4000"]
    metrics_path: /metrics