Run started 2026-10-18 10:38 UTC
· completed 10:46 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
63.10s
p50
59.66s
p90
68.49s
p99
114.84s
min
42.92s
max
119.99s
std
20.52s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
5.34s
p50
5.01s
p90
7.82s
p99
9.41s
min
2.67s
max
9.59s
std
2.01s
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
78.69s
p50
76.56s
p90
89.09s
p99
121.62s
min
52.10s
max
125.24s
std
18.73s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
43.03s
p50
44.55s
p90
58.28s
p99
60.71s
min
20.55s
max
60.98s
std
13.29s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
24.23s
p50
13.07s
p90
48.76s
p99
63.69s
min
10.25s
max
65.35s
std
18.71s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.38s
p50
0.00s
p90
0.82s
p99
3.14s
min
0.00s
max
3.40s
std
1.03s
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
68.2%
p50
75.0%
p90
87.2%
p99
89.3%
min
29.8%
max
89.6%
std
20.1%
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
35.3%
p50
26.7%
p90
68.7%
p99
70.5%
min
14.7%
max
70.7%
std
20.3%
Tool Calls
Number of tool spans the agent produced for the case.
mean
11.09
p50
11.00
p90
13.00
p99
15.70
min
7.00
max
16.00
std
2.47
sum
122.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).