Run started 2026-10-17 09:27 UTC
· completed 09:36 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
103.95s
p50
104.94s
p90
131.65s
p99
141.82s
min
66.81s
max
142.95s
std
27.07s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
13.25s
p50
13.04s
p90
18.34s
p99
19.66s
min
7.07s
max
19.81s
std
4.56s
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
129.46s
p50
122.23s
p90
180.76s
p99
288.08s
min
0.00s
max
300.00s
std
77.14s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
77.66s
p50
74.56s
p90
100.87s
p99
113.29s
min
52.40s
max
114.67s
std
20.86s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
28.82s
p50
23.77s
p90
53.41s
p99
58.67s
min
8.73s
max
59.25s
std
19.27s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
2.49s
p50
1.29s
p90
5.74s
p99
9.43s
min
0.00s
max
9.84s
std
3.43s
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
73.4%
p50
76.5%
p90
86.5%
p99
90.6%
min
53.9%
max
91.0%
std
13.2%
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
26.0%
p50
23.9%
p90
41.0%
p99
41.9%
min
6.9%
max
42.0%
std
13.8%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.12
p50
12.00
p90
15.00
p99
15.00
min
9.00
max
15.00
std
2.42
sum
97.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).