Usage
Monitor consumption, token usage, and costs across your organization.
Dashboard usage page
Open Usage from the dashboard sidebar (/usage). The page shows:
- Daily request counts (completed and failed)
- Token usage and costs over time
- Breakdowns by engine, model, and provider
- Recent generation requests with status and latency
API usage endpoint
Query usage programmatically with GET /v1/usage:
curl https://api.pipevideo.co/v1/usage \
-H "Authorization: Bearer pv_your_api_key"The response includes:
| Field | Description |
|---|---|
credit_balance | Current balance in cents |
totals | Aggregate request counts and total cost |
daily_usage | Per-day request and cost breakdown |
daily_usage_by_route | Per-day breakdown by engine, model, and provider |
recent_requests | Latest generations with status, cost, and latency |
See Usage API for the full response schema.
What drives usage
Usage scales with:
- Number of generation requests — each
POST /v1/responsescreates one generation - Orchestration token consumption — input and output tokens from the LLM composition step
- Model choice — different orchestration models have different per-token rates
Rendering (HyperFrames) is included in the orchestration billing model — there is no separate render fee.
Monitoring recommendations
- Set up internal alerts when
credit_balancedrops below your team's threshold (pollGET /v1/usageor check the dashboard). - Review
daily_usage_by_routeto identify high-cost model or engine combinations. - Use webhooks in production to avoid unnecessary polling overhead.