Run started 2026-10-18 07:24 UTC
· completed 07:32 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
100.80s
p50
78.73s
p90
187.77s
p99
190.60s
min
49.41s
max
190.91s
std
51.84s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
10.48s
p50
9.44s
p90
13.43s
p99
18.88s
min
5.21s
max
19.49s
std
3.88s
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
104.67s
p50
97.89s
p90
181.50s
p99
198.34s
min
40.75s
max
198.43s
std
46.59s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
64.36s
p50
55.41s
p90
115.11s
p99
131.24s
min
23.75s
max
133.03s
std
33.91s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
36.77s
p50
30.46s
p90
63.93s
p99
70.57s
min
14.64s
max
71.31s
std
18.84s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
2.59s
p50
2.39s
p90
4.66s
p99
7.90s
min
0.00s
max
8.26s
std
2.50s
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
62.7%
p50
64.0%
p90
73.1%
p99
73.2%
min
46.3%
max
73.2%
std
9.2%
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
37.3%
p50
33.6%
p90
55.0%
p99
61.4%
min
23.1%
max
62.1%
std
13.0%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.40
p50
9.50
p90
20.10
p99
20.91
min
8.00
max
21.00
std
5.02
sum
124.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).