Run started 2026-10-18 09:24 UTC
· completed 09:35 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
83.02s
p50
79.29s
p90
116.96s
p99
129.63s
min
38.91s
max
131.31s
std
27.60s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
5.63s
p50
5.05s
p90
6.79s
p99
12.01s
min
3.37s
max
12.71s
std
2.41s
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
95.32s
p50
94.98s
p90
130.42s
p99
138.95s
min
60.30s
max
139.86s
std
26.05s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
51.83s
p50
46.05s
p90
74.05s
p99
80.97s
min
34.95s
max
81.58s
std
15.38s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
38.04s
p50
33.07s
p90
60.99s
p99
84.02s
min
7.14s
max
86.92s
std
21.30s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.50s
p50
0.00s
p90
1.57s
p99
2.56s
min
0.00s
max
2.68s
std
0.85s
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
62.5%
p50
65.8%
p90
77.4%
p99
83.9%
min
40.3%
max
84.6%
std
13.9%
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
43.3%
p50
43.4%
p90
64.0%
p99
71.8%
min
16.2%
max
72.8%
std
17.9%
Tool Calls
Number of tool spans the agent produced for the case.
mean
11.08
p50
10.00
p90
13.00
p99
15.64
min
9.00
max
16.00
std
2.18
sum
144.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).