Sign InOpen Brain
VercelEngineering PostOfficial Source

AI Gateway logs now have a dedicated page

AI Gateway’s dedicated logs expose per-request cost, tokens, latency, routing, and provider fallbacks, making agent failures and spend anomalies easier to trace.

Vercel · Jul 31, 2026
Open Source Open MarkdownOpen JSON
Source Summary

The new **Logs page** lists gateway requests with cost, token counts, duration, model, provider, and region. Request details include time to first token, cache and reasoning tokens, policy flags, and the ordered fallback path.

Practical Implication

Use filters and request-ID search to debug agent failures or compare routing behavior. Filtered views are shareable by URL and exportable as **CSV or JSON**, which makes targeted cost and reliability analysis easier.

Agent-Ready Context
The new **Logs page** lists gateway requests with cost, token counts, duration, model, provider, and region. Request details include time to first token, cache and reasoning tokens, policy flags, and the ordered fallback path.

Use filters and request-ID search to debug agent failures or compare routing behavior. Filtered views are shareable by URL and exportable as **CSV or JSON**, which makes targeted cost and reliability analysis easier.

The page records what happened but does not define alerts, retention, or automated remediation. Diagnosis still requires interpreting provider failures, routing-budget exhaustion, and latency in the context of the agent run.
Connected Context · Feed7 Judgment

This closes a diagnostic gap around gateway routing by exposing the actual provider path, policy decisions, latency components, and token economics per request. It complements existing routing and cost controls with inspectable evidence, but remains a human analysis surface rather than an alerting or remediation system. Export and shareable filters make the same evidence usable in targeted evaluations.

AI Gateway now supports team and project spend budgetsBudgets enforce and alert on aggregate spending, while the logs provide request-level cost and token evidence needed to explain which routes consumed those budgets.AI Gateway adds unified fast mode supportFast mode is only a preference with possible standard-speed fallback; route, duration, time-to-first-token, and pricing metadata in the logs let teams verify what actually happened.Claude Fable 5 access restored on AI GatewayThat release recommends fallbacks because stricter classifiers may refuse coding calls; the ordered fallback path and policy flags now provide evidence for diagnosing those routing outcomes.From Signal to PR: Anatomy of a Self-Improving Agent — Jason Lopatecki, ArizeThe self-improving-agent workflow depends on production traces and observability; gateway logs add routing, latency, policy, and cost evidence that can feed such investigations, without automating the improvement loop itself.
Context Map
infracodingdata#observability#gateways#agent-reliability
Uncertainty
The page records what happened but does not define alerts, retention, or automated remediation. Diagnosis still requires interpreting provider failures, routing-budget exhaustion, and latency in the context of the agent run.