AI Ops
Usage, latency, cost and quality for every model and agent.
Requests
1.28M
18.2% vs last month
Tokens
168.4M
21.6% vs last month
p95 latency
842 ms
4.1% vs last month
Spend
$4,210
6.3% vs last month
Token usage
Last 14 days, in millions
Monthly budget
80% used$3,380 / $4,200
- support-agent$1,932
- search-rag$924
- doc-extractor$524
At the current rate you'll reach the budget on Nov 6. Raise the limit
Latency
Today, by hour
Requests by model
4
models
- nova-large48%
- nova-mini31%
- embed-v314%
- vision-27%
Agents
Manage| Agent | Model | Requests | Eval score | Error rate |
|---|---|---|---|---|
support-agent v4.2 |
nova-large | 612,400 | 94 |
0.4% |
search-rag v2.8 |
nova-mini | 388,210 | 89 |
0.7% |
doc-extractor v1.5 |
vision-2 | 142,880 | 91 |
1.2% |
sales-copilot v0.9 |
nova-large | 96,300 | 78 |
2.6% |
triage-bot v3.1 |
nova-mini | 40,210 | 86 |
0.3% |
Live traces
Streaming-
1.2s
Where is my refund for order 10472?
support-agent
-
0.6s
Find the SOC 2 report from 2025
search-rag
-
3.8s
Parse invoice PDF (4 pages)
doc-extractor
-
2.1s
Draft follow-up for Kestrel Bank
sales-copilot
-
0.9s
Cancel my subscription
support-agent