Observability

Spend tracking

See tokens, cost and spend per request, model, provider, key and project.

Cost per requestInput and output tokensCost limits and alerts

What is Spend tracking?

Spend tracking is how an AI gateway records the tokens, cost and price of every request, so you can see spend by model, provider, API key and project. On Spro each request gets a usage log row with input and output tokens, cost, latency and the key used. A monthly cost limit guardrail and balance alerts help you stay within budget.

LLM bills are hard to predict. One prompt change or a runaway agent can multiply tokens without warning. Spend tracking turns every request into a measurable row, so you can find the expensive models, keys and features.

Spro records tokens and cost for each request in the usage log, including BYOK requests. BYOK requests are recorded for visibility but are not charged against your Spro credit balance, because your provider bills them.

Key features

What this feature does for your requests.

  • Input and output tokens recorded for every request
  • Cost per request, based on the price of the model that served it
  • Spend grouped by model, provider, API key, project and trace
  • A monthly cost limit guardrail that restricts activity when reached
  • Alerts for low balance, balance depletion and scheduled spend reports

How it works

4 steps, from setup to the first request.

  1. 01

    Send requests

    Make requests with a Spro key, from any project.

  2. 02

    Open usage

    Go to the workspace usage page to see tokens, cost and latency.

  3. 03

    Filter

    Group or filter by model, key or trace to find the cost driver.

  4. 04

    Set limits

    Add a cost limit guardrail or a balance alert so spend does not surprise you.

Is it the right fit?

Specific situations where this feature helps, and where it does not.

Good fit if

  • Several products or teams share one AI budget and need to see who spends what
  • You want to find the model or feature behind a sudden cost jump
  • Finance needs numbers that trace back to individual requests

Not a good fit if

  • You only need a monthly total from your provider invoice
  • You need billing-exact figures. Costs use listed model prices, so check invoices for the official amount

Spend tracking FAQs

Common questions about Spend tracking.

Route every AI request through SPRO.

One OpenAI-compatible API for 350+ models, with intelligent routing, 29 built-in guardrails, observability, and automatic failover.

Talk to sales