MagicDash DocsMagicAssistant
Analytics
Track AI usage, costs, and performance
Analytics & Logging
Monitor your AI usage. Track API consumption, costs, response times, and errors to optimize performance and manage spending.
Dashboard Overview
Access at MagicAssistant → Analytics:
- Total Requests — API calls made
- Total Tokens — Input + output tokens consumed
- Total Cost — Estimated spending
- Success Rate — Percentage of successful requests
Cost Monitoring
Track spending across providers:
- By Provider — OpenAI, Anthropic, Google, etc.
- By Model — GPT-4, Claude, Gemini, etc.
- By Feature — Chat, images, SEO, etc.
- By Day — Daily spending chart
Set budget alerts to notify you when approaching limits.
Response Times
Monitor AI performance:
- Average — Typical response time
- P95 — Worst-case typical
- Trends — Performance over time
Error Tracking
Understand API failures:
| Error | Meaning |
|---|---|
| Rate Limited | Too many requests |
| Invalid API Key | Key expired or wrong |
| Insufficient AI Credits | Out of AI Credits |
| Context Too Long | Conversation too big |
Date Filtering
View data for:
- Today, yesterday, last 7/30 days
- This month, last month
- Custom date ranges
- Compare two periods
Export Data
Export in CSV, JSON, or PDF for external analysis.