Run started 2026-10-19 08:05 UTC
· completed 08:44 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
101.13s
p50
95.26s
p90
155.67s
p99
197.48s
min
22.12s
max
199.32s
std
44.46s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
10.52s
p50
9.53s
p90
19.12s
p99
26.52s
min
2.85s
max
28.59s
std
5.77s
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
114.03s
p50
113.10s
p90
173.64s
p99
210.43s
min
28.55s
max
213.95s
std
46.64s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
71.35s
p50
64.23s
p90
119.10s
p99
141.93s
min
16.33s
max
147.53s
std
32.75s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
34.70s
p50
32.32s
p90
53.08s
p99
102.46s
min
8.29s
max
106.54s
std
21.51s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
1.00s
p50
0.00s
p90
2.88s
p99
7.83s
min
0.00s
max
8.10s
std
2.07s
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
68.6%
p50
68.9%
p90
80.3%
p99
83.3%
min
45.6%
max
84.2%
std
10.0%
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
33.1%
p50
29.5%
p90
51.2%
p99
68.4%
min
15.8%
max
71.4%
std
11.8%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.41
p50
12.00
p90
19.00
p99
23.62
min
4.00
max
24.00
std
4.11
sum
484.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).