Run started 2026-10-19 06:05 UTC
· completed 07:02 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
71.86s
p50
64.42s
p90
117.42s
p99
196.02s
min
16.88s
max
220.35s
std
37.68s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
10.31s
p50
7.26s
p90
17.59s
p99
45.93s
min
3.66s
max
49.31s
std
8.91s
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
95.62s
p50
87.64s
p90
139.22s
p99
239.47s
min
29.28s
max
240.38s
std
43.25s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
61.98s
p50
50.51s
p90
108.40s
p99
170.91s
min
14.85s
max
205.13s
std
35.29s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
13.07s
p50
10.01s
p90
24.49s
p99
50.65s
min
2.05s
max
57.36s
std
10.23s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.94s
p50
0.20s
p90
1.98s
p99
7.09s
min
0.00s
max
25.59s
std
2.65s
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
82.2%
p50
84.7%
p90
94.2%
p99
117.3%
min
38.6%
max
123.8%
std
13.3%
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
18.2%
p50
15.9%
p90
32.7%
p99
49.0%
min
4.0%
max
58.8%
std
10.9%
Tool Calls
Number of tool spans the agent produced for the case.
mean
11.45
p50
11.00
p90
15.00
p99
19.96
min
7.00
max
23.00
std
2.86
sum
1,202.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).