Usage & Billing

Usage Tracking

Monitor your HyperMemory usage

Usage Tracking

Track your node count and query usage to stay within plan limits and avoid unexpected charges.

Usage dashboard

Accessing the dashboard

  1. Go to app.hypermemory.io
  2. Navigate to SettingsUsage

Current period overview

The dashboard shows:

MetricDescription
NodesCurrent node count / plan limit
QueriesQueries used this month / plan limit
Billing periodCurrent month dates
Days remainingDays until usage resets

Visual indicators

  • Green — Under 70% of limit
  • Yellow — 70-90% of limit
  • Red — Over 90% of limit

Historical usage

Charts

View usage trends over time:

  • Nodes over time — Growth of your knowledge base
  • Daily queries — Query volume per day
  • Queries by API key — Which keys are most active

Export data

Export usage data for analysis:

  1. Click Export on the Usage page
  2. Select date range
  3. Choose format (CSV or JSON)
  4. Download

Usage by graph

If you have multiple graphs, see usage breakdown:

GraphNodes% of Total
production45,23068%
staging18,44228%
testing2,8914%

Real-time monitoring

API endpoint

Query your usage programmatically:

import httpx

response = httpx.get(
    "https://api.hypermemory.io/v1/usage",
    headers={"Authorization": f"Bearer {api_key}"}
)

usage = response.json()
print(f"Nodes: {usage['nodes']['current']}/{usage['nodes']['limit']}")
print(f"Queries: {usage['queries']['current']}/{usage['queries']['limit']}")

Response:

{
  "nodes": {
    "current": 45230,
    "limit": 100000,
    "percentage": 45.2
  },
  "queries": {
    "current": 127543,
    "limit": 250000,
    "percentage": 51.0
  },
  "billing_period": {
    "start": "2026-03-01T00:00:00Z",
    "end": "2026-03-31T23:59:59Z"
  }
}

Webhooks

Get notified when usage thresholds are crossed:

  1. Go to SettingsWebhooks
  2. Add a webhook URL
  3. Select events:
    • usage.nodes.threshold (70%, 90%, 100%)
    • usage.queries.threshold (70%, 90%, 100%)

Webhook payload:

{
  "event": "usage.queries.threshold",
  "threshold": 90,
  "current": 225000,
  "limit": 250000,
  "timestamp": "2026-03-25T14:30:00Z"
}

Usage alerts

Email alerts

Configure email notifications:

  1. Go to SettingsAlerts
  2. Enable usage alerts
  3. Set thresholds:
    • Warning — e.g., 70% of limit
    • Critical — e.g., 90% of limit

Slack integration

Send alerts to Slack:

  1. Go to SettingsIntegrations
  2. Connect Slack workspace
  3. Select channel for alerts
  4. Configure alert types

Understanding your usage patterns

Query distribution

Analyze when queries happen:

import httpx

response = httpx.get(
    "https://api.hypermemory.io/v1/usage/queries/distribution",
    params={"period": "day"},
    headers={"Authorization": f"Bearer {api_key}"}
)

for hour, count in response.json()["distribution"].items():
    print(f"{hour}:00 - {count} queries")

Top operations

See which memory operations are most used:

OperationCount% of Queries
memory_recall89,23470%
memory_find_related25,42120%
memory_get_relationships12,88810%

By API key

Identify which keys drive usage:

API KeyQueries% of Total
prod-agent102,54380%
staging-test20,00016%
ci-pipeline5,0004%

Optimization tips

Reduce query volume

Cache frequent queries — If your agent repeatedly asks the same question, cache the result locally.

Use specific queries — “What is Sarah’s communication preference?” is more efficient than “Tell me everything about Sarah.”

Batch operations — Instead of 10 separate recalls, use one recall with broader scope and filter client-side.

Manage node growth

Archive old nodes — Use metadata to mark nodes as archived instead of keeping them active.

Consolidate redundant nodes — Merge duplicate or overlapping memories.

Clean up test data — Regularly purge test nodes from development graphs.

Billing period

  • Start: First day of calendar month (00:00 UTC)
  • End: Last day of calendar month (23:59 UTC)
  • Reset: Query counts reset to zero on the first

Node counts don’t reset — they reflect your current storage. Only query counts reset monthly.

Next steps