Cloudflare AI Gateway analytics
Cloudflare AI Gateway analytics shows requests, tokens, cost, latency, and errors across all providers connected to your gateway.
Hub: Tokenminning with Cloudflare AI Gateway.
This page answers cloudflare ai gateway analytics, cloudflare ai gateway cost, and cloudflare ai gateway monitoring.
Open analytics
Cloudflare dashboard → AI Gateway → your gateway → Analytics:
| Metric | Use |
|---|---|
| Requests | Volume trend — spike detection |
| Tokens | Input vs output ratio |
| Cost | Provider spend through gateway |
| Errors | Failed fallbacks, rate limits |
| Cache hit rate | Caching effectiveness |
See Analytics docs .
Weekly review checklist
- Which model slug dominates cost?
- Did request volume spike on any route?
- Is cache hit rate climbing after enabling caching?
- Any error rate increase from fallback chains?
Logging
Enable logging for request-level detail. Logs complement Analytics for debugging one expensive call.
Cost attribution
Use separate gateway routes or IDs per feature:
/gateway/prod-chat → customer-facing
/gateway/internal-agent → engineering agents
/gateway/ci-eval → CI scriptsFilter Analytics by gateway route for cost attribution.
Spending guardrails
- Review Analytics weekly — top 3 models by cost
- Alert when daily cost exceeds threshold
- Compare cache hit rate before/after prompt stabilization
- Trace errors before adding more fallback targets (each retry costs tokens)
Related
Last updated on