Opsmeter.io
AI Cost & Inference Control
Integration docs
Implementation guide for AI cost control
Ship telemetry fast, keep attribution stable, and move from ingest to governance without changing your app network path.
Recommended path
Start with direct ingest, validate one stable payload shape, then harden budgets, retries, and incident workflows.
What this page answers
- What happens when a plan limit pauses telemetry ingest.
- How budget warning and budget exceeded states are calculated.
- Why provider calls continue even when Opsmeter.io telemetry pauses.
Implementation rhythm
5 minfirst ingest path
No proxyapp path unchanged
One schemaacross providers
Use quickstart for initial payloads, then operations docs to harden production workflows.
Limits and budgets
Plan limit and budget are different controls. Plan limit pauses telemetry ingest when request capacity is exhausted. Budget tracks spend thresholds (daily and monthly) for warning/exceeded workflows.
| Control | What it monitors | When it triggers | Effect |
|---|---|---|---|
| Plan limit | Telemetry request count | Remaining request count reaches 0 | Telemetry paused (402, reason plan_limit_reached) |
| Budget warning | Spend utilization ratio | Daily or monthly spend reaches warning threshold | Warning only, telemetry still active |
| Budget exceeded | Spend versus configured limit | Daily or monthly spend goes above limit | Exceeded signal, hard-stop depends on workspace policy |
Workspace status endpoint
Use GET /v1/settings/workspace-status (Bearer) or GET /v1/diagnostics/workspace-status (X-API-Key) to read current state with top-level booleans and nested daily/monthly budget windows.
{
"workspaceId": "f0f2d17d-7bf7-4efa-a11e-4f13e8a0216b",
"isPlanLimitReached": false,
"telemetryPaused": false,
"providerCallsContinue": true,
"isBudgetWarning": true,
"isBudgetExceeded": false,
"severity": "warning",
"plan": {
"planTier": "starter",
"planName": "Starter",
"currentRequestCount": 98231,
"planRequestLimit": 100000,
"remainingRequestCount": 1769,
"isPlanLimitReached": false,
"telemetryPaused": false,
"periodStart": "2026-03-01T00:00:00Z",
"periodEnd": "2026-03-31T23:59:59Z"
},
"budget": {
"warningThresholdRatio": 0.8,
"daily": {
"limitUsd": 50,
"currentSpendUsd": 41.25,
"utilizationRatio": 0.825,
"isWarning": true,
"isExceeded": false,
"period": "daily",
"periodStart": "2026-03-07T00:00:00Z",
"periodEnd": "2026-03-07T23:59:59Z"
},
"monthly": {
"limitUsd": 500,
"currentSpendUsd": 412.55,
"utilizationRatio": 0.8251,
"isWarning": true,
"isExceeded": false,
"period": "monthly",
"periodStart": "2026-03-01T00:00:00Z",
"periodEnd": "2026-03-31T23:59:59Z"
}
}
}Plan-limit blocked ingest response
When telemetry ingest is blocked by plan capacity, ingest returns 402 with explicit fields.
{
"ok": false,
"accepted": false,
"reason": "plan_limit_reached",
"telemetryPaused": true,
"providerCallsContinue": true,
"isPlanLimitReached": true,
"budgetWarning": false,
"budgetExceeded": false,
"planLimitRequests": 100000,
"freeLimitRemainingRequests": 0
}Telemetry pause affects Opsmeter.io ingest only. Your provider calls continue normally.