AI operations
Inference health, token economics and execution signals from one demo dataset.
Inference operations
Demo telemetryTotal API requests
42
92.9% successful · filtered request records
Successful
39
Error rate
7.1%
Average latency
1136 ms
Tokens consumed
89,051
Estimated cost
$0.2412
Token & cost intelligence
Striped: input · Solid: output. Sample token counts, including input on errors.
Model performance leaderboard
Ranked by mean latencyOrzen Flash
#1 · 1.7 requests / day
Orzen Core
#2 · 2.6 requests / day
Atlas Reason
#3 · 1.7 requests / day
Recent AI requests
Inspect all →Execution & prompt activity
Workflows →Runs in scope
2
Active
0
Failed
1
Operational guardrails
2 signals to reviewReview scoped operational signals
Demo checks only · no real limits or alerts are enforced
Estimated budget
$0.2412
$1.7588 remaining · 12% used
Within allowanceToken allowance
89,051
210,949 remaining · 30% used
Within allowanceUsage distribution
18 requests · 37,569 tokens · $0.0835
12 requests · 25,442 tokens · $0.0072
12 requests · 26,040 tokens · $0.1504
Workspace request share
Latency profile
All scoped requests · millisecondsP50 / median
680ms
P95
2,825ms
P99
2,975ms
Select a latency bucket to inspect its request outcomes.
Request source intelligence
Origin & environment28 successful · 2 failed · 1162 ms mean
11 successful · 1 failed · 1070 ms mean
0 successful · 0 failed · 0 ms mean
Production sample records
30
Sandbox sample records
12