Skip to main content

Overview

AI Usage is a single-page analytics dashboard showing usage costs, call volumes, duration, and profit for your agents across every connected AI provider. Everything — period filter, provider tabs, headline metrics, cost pies, top-agent chart, and per-agent tiles — lives on the same page; no modal hop required to read the basics. AI Usage Analytics page — period filter, provider tabs, 5 headline metrics, and cost / provider / success-rate pie charts

Where to find it

  1. Partner Portal → Agent Analytics in the sidebar. The page header reads “AI Usage Analytics” — the sidebar and header use slightly different names for the same page.

Supported providers

Voice-agent providers (matching the Provider filter row below):
  • VAPI Agents
  • Retell Agents
  • Ultravox Agents
  • GHL Agents
  • ElevenLabs Agents
Chat-agent providers (filter row only):
  • Retell Chat
Provider list vs filter row: Knova Agents is a live voice-agent provider, but may not appear on the AI Usage filter row. If you don’t see Knova calls on this page, check the per-agent View Detailed Analytics modal — Knova spend appears there even when missing from the filter row.

Metrics tracked

MetricDescription
Total CallsNumber of calls made
Failed CallsCalls that ended with a failed status
Success RatePercentage of non-failed calls
Average DurationAverage call length
Total CostActual cost from the provider
Avg Cost/CallAverage cost per call
Billed CostTotal cost × profit multiplier
ProfitBilled Cost minus Actual Cost

Filter by period and provider

The top of the page has two filter rows:
  • Period: Day / Week / Month — recalculates every metric and chart for the selected window.
  • Provider: All / VAPI / RETELL / ULTRAVOX / GHL / ElevenLabs / Retell Chat — narrows the page to one provider’s agents (or All for cross-provider totals).
Changing either filter triggers a refresh of headline metrics, pie charts, the Top Agents by Cost chart, and the agent grid below.

Cost breakdown

Three pie charts sit directly below the headline metrics on the same page:
  • Product Cost Breakdown — share of cost across product types (LLM / STT / TTS / Provider fees)
  • Cost by Provider — share of total spend by upstream provider
  • Call Success Rate — successful vs failed call ratio
For a single-agent breakdown of LLM / STT / TTS / Provider costs, open the agent’s View Detailed Analytics modal from the agent grid (see below).

Agent tiles and detailed analytics

Below the pies, the page renders a Top Agents by Cost bar chart, then a grid of per-agent tiles. Each tile shows the agent’s headline numbers (Total Calls, Billable Cost, Actual Cost, Profit, Avg Duration, Success Rate, Cost/Min) and a View Detailed Analytics button that opens a deeper modal for that single agent. Top Agents by Cost chart and AI Agents grid with View Detailed Analytics buttons
The provider badge on each tile (RETELL, ElevenLabs, etc.) reflects the upstream provider; tiles from disabled providers won’t appear under the All filter.

Profit analysis

The analytics modal shows a profit analysis bar chart:
  • Actual Cost — what you pay the provider
  • Billed Cost — what you charge the customer
  • Profit — difference between billed and actual

Retell summary

The Retell tab shows a summary row at the top with:
  • Total Cost
  • Total Duration
  • Total Calls
  • Success Rate
Plus a cost breakdown grid of product costs before individual agent tiles.

Refresh data

Click Refresh Summary to reload aggregate data for the Retell tab.

API key issues

If an API key is missing for a provider, you’ll see a red error banner with a button to navigate to Settings to configure the key.

Next steps