Use Case — Cost Control
See exactly where your LLM budget is going — before the invoice arrives.
LLM costs compound fast. One agent running GPT-4o at high volume can exceed a monthly budget in days. Zespan gives you attribution, forecasting, and AI recommendations in one place.
The problem
The cloud bill can't tell you who spent what
OpenAI invoices show spend by model, not by agent, feature, or user. You know GPT-4o is expensive — not which part of your product is calling it.
Cost spikes show up a month late
A new agent launched two weeks ago and has been hammering GPT-4o. You find out when the monthly invoice arrives.
You know it's high — not why
Total spend is too high but you don't know whether to switch models, compress prompts, add caching, or fix a runaway agent.
How to use Zespan for this
Open Cost Attribution — slice spend any way you need
In the sidebar, open Cost → Attribution. The view shows total LLM spend broken down by agent, model, user, operation, or tool. Switch the dimension selector to find your most expensive agent, the costliest model, or the operation type driving the most tokens. Time ranges: 24h, 7d, 30d, 90d.
Open Cost Explorer — see the daily trend
Cost Explorer shows your daily spend as a time-series chart. Scroll back to when a new agent launched. You'll see exactly when the step-change happened. Anomaly detection markers flag days where spend deviated significantly from the trend — catching spikes the day they happen, not at month end.
Check the Forecast — know where you're heading
Open Cost → Forecast. Zespan projects spend 30 days forward using regression over your daily history, with a confidence band showing best/worst case. If the trend line is heading toward a budget breach, you see the number now — with time to act before it happens.
Read AI Recommendations — ranked cuts with savings estimates
Open Cost → Recommendations. Zespan analyzes your trace data and generates specific, ranked suggestions: 'Switch agent X from GPT-4o to GPT-4o-mini — saves $420/month' or 'Compress the summarization prompt — removes 40% boilerplate'. Each recommendation has a projected monthly savings, confidence level, and a copy-paste code snippet.
Set a cost alert — get paged before overage
In Alerts, create a rule: total_cost > $50 in a 60-minute window, fire to Slack. Now a cost spike triggers a page instead of a surprise invoice. Pair with a weekly budget review in ZespanPilot: ask 'What was my spend breakdown by agent this week?' and get a plain-English summary.
Start free — 10K traces/month, no card needed
See every agent decision, tool call, and handoff in production. Setup takes under 5 minutes.
Get started free →