Run started 2026-10-17 09:08 UTC
· completed 09:15 UTC
Headline Metrics
A case passes when (Correctness or String in Response) and Relevance and Tools pass.
Scorers that had nothing to judge are ignored; cases where no scorer produced a result are
counted as no data and excluded from every pass rate.
First Token Time
Time to the first response token (TTFT) — how long before the user sees anything.
mean
80.08s
p50
82.01s
p90
111.26s
p99
118.39s
min
44.30s
max
119.90s
std
26.57s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
9.32s
p50
8.76s
p90
13.34s
p99
14.64s
min
5.21s
max
14.70s
std
2.83s
Case Duration
Wall-clock time the eval harness measured for the whole case, including scoring and gRPC overhead. Always at least the agent's own time_total.
mean
92.11s
p50
91.29s
p90
123.26s
p99
142.74s
min
51.38s
max
146.77s
std
28.95s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
63.93s
p50
60.01s
p90
90.08s
p99
106.78s
min
32.95s
max
108.84s
std
22.70s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
19.17s
p50
19.63s
p90
30.04s
p99
36.77s
min
4.14s
max
38.08s
std
8.72s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.76s
p50
0.06s
p90
2.58s
p99
3.45s
min
0.00s
max
3.48s
std
1.15s
LLM Time Share
LLM time as a percentage of the agent's total span duration. This is the spread of per-case percentages, not the run-level ratio.
mean
77.6%
p50
77.3%
p90
90.9%
p99
92.5%
min
59.9%
max
92.5%
std
9.8%
Tool Time Share
Tool time as a percentage of the agent's total span duration. This is the spread of per-case percentages, not the run-level ratio.
mean
23.5%
p50
21.9%
p90
34.8%
p99
36.4%
min
7.7%
max
36.5%
std
8.5%
Tool Calls
Number of tool spans the agent produced for the case.
mean
10.79
p50
10.00
p90
13.20
p99
14.82
min
6.00
max
15.00
std
2.35
sum
205.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).