LLM calls
...
Last 7d
query_stats
LLM Usage Analytics
Calls, token volume, spend, latency, and failures across recommendations, roadmap generation, chat, embeddings, scoring, and reindex workflows.
Total tokens
...
Input + output where reported
Estimated spend
...
Provider usage metadata
Error rate
...
0 failed calls
P95 latency
...
Successful and failed calls
Cached input
...
Reusable prompt tokens
Usage Over Time
Loading window
No LLM events were recorded in this window.
Feature Load
Top feature: No feature activity
No breakdown data available.
Model Mix
Token and cost concentration by provider model.
No model usage data available.
Operation Hotspots
Highest usage: No operation activity
No breakdown data available.
Features
| Name | Calls | Tokens | Cost | Error rate | Avg latency |
|---|---|---|---|---|---|
| No usage rows in the selected window. | |||||
Operations
| Name | Calls | Tokens | Cost | Error rate | Avg latency |
|---|---|---|---|---|---|
| No usage rows in the selected window. | |||||
Recent LLM Events
Latest tracked calls with status, tokens, latency, cost, and linked student context.
| Time | Feature | Operation | Model | Status | Tokens | Latency | Cost | Student | Error |
|---|---|---|---|---|---|---|---|---|---|
| No recent LLM events in the selected window. | |||||||||