Unified Reporting for All AI Gateway Usage

AI Gateway's Custom Reporting API has been released in beta for Pro and Enterprise plans. Now you can query cost, token usage, and request volume scattered across multiple providers, keys, and dashboards from a single endpoint.

You can break down costs by model, provider, user ID, custom tag, and credential type, including BYOK requests. This lets you treat usage and spend by feature, customer, and plan as operational data, and you can also query in real time with Claude Code.

As an example, one AI platform handling 200K+ users eliminated a separate proxy layer during the private beta period and combined cost tracking with request management into one, saving $80K+. Since then, they've been using Advanced Reporting along with custom tags and user IDs to track usage and cost per customer across models.

The key is attaching user and tags to every request so that product and finance teams can read costs in units they understand.

• Tag requests by customer ID, plan, and feature
• Supports AI SDK, Chat Completions API, Responses API, OpenResponses API, and Anthropic Messages API
• Aggregates both BYOK and system credentials at the same reporting endpoint
• Query periodically to inform budget, margin, and pricing decisions

AI spend now becomes an operational metric rather than after-the-fact reconciliation. By consolidating traffic through a single reporting endpoint, you can catch cost spikes before the invoice arrives and judge per-feature unit economics and profitability with real data.