Usage tracking
The 诺玛AI console provides comprehensive usage statistics and performance monitoring.
Real-time usage statistics#
Overview panel
The console home page shows key metrics:
- Total requests — total number of API calls in the current period
- Total tokens — total prompt + completion tokens consumed
- Total cost — accumulated cost in the current period
- Success rate — the API request success rate
Time trends
You can view trends across different time dimensions:
Dimension
Description
Real-time
Last 1 hour, per-minute granularity
Day
Last 24 hours, per-hour granularity
Week
Last 7 days, per-day granularity
Month
Last 30 days, per-day granularity
Multi-dimensional analysis#
Analysis by model
View the usage of each model:
- Share of request count
- Share of token consumption
- Share of cost
- Average latency
Analysis by provider
View the performance of each provider node:
- Success rate comparison
- Latency distribution
- Fallback trigger frequency
Analysis by API Key
View the usage of each API Key:
- Request volume per key
- Cost per key
- Makes it easy to track cost by project/team
Anomaly alerts#
Alert rules
You can configure the following alerts:
Alert type
Description
Error rate alert
Error rate exceeds a threshold (e.g. 5%)
Cost alert
Daily/weekly/monthly cost exceeds budget
Latency alert
Average latency exceeds a threshold
Fallback alert
Abnormal fallback trigger frequency
Notification channels
Alerts support the following notification methods:
- Email notifications
- Webhook (can integrate with Slack, Feishu, DingTalk, etc.)
Alert configuration is managed in the console under Settings → Alert rules.
Best practices#
- Use a different API Key for each project — makes it easy to track cost by project
- Set cost alerts — avoid unexpected large bills
- Review the model distribution regularly — ensure model selection matches expectations
- Watch the error rate trend — an abnormal rise may indicate a configuration issue
Last updated on June 23, 2026