Run started 2026-10-18 11:28 UTC
· completed 11:42 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
84.93s
p50
78.87s
p90
116.29s
p99
151.56s
min
43.97s
max
155.48s
std
30.91s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
8.95s
p50
6.09s
p90
14.05s
p99
26.15s
min
3.73s
max
27.49s
std
7.11s
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
93.91s
p50
85.90s
p90
118.66s
p99
160.93s
min
53.80s
max
165.62s
std
30.66s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
56.93s
p50
50.70s
p90
73.63s
p99
110.70s
min
35.72s
max
114.82s
std
22.88s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
34.06s
p50
33.42s
p90
49.23s
p99
62.55s
min
10.25s
max
64.03s
std
16.68s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.00s
p50
0.00s
p90
0.00s
p99
0.00s
min
0.00s
max
0.00s
std
0.00s
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
67.0%
p50
68.7%
p90
81.7%
p99
88.8%
min
44.6%
max
89.6%
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
38.5%
p50
35.0%
p90
56.7%
p99
57.8%
min
17.9%
max
58.0%
std
14.0%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.90
p50
13.00
p90
16.20
p99
17.82
min
9.00
max
18.00
std
2.96
sum
129.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).