Run started 2026-10-18 09:59 UTC
· completed 10:18 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
91.63s
p50
91.47s
p90
114.52s
p99
117.18s
min
57.42s
max
117.59s
std
19.42s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
11.82s
p50
9.10s
p90
26.15s
p99
39.96s
min
2.64s
max
41.50s
std
10.98s
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
104.26s
p50
106.63s
p90
122.97s
p99
129.31s
min
63.17s
max
130.36s
std
18.96s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
58.75s
p50
59.87s
p90
80.56s
p99
82.64s
min
40.65s
max
82.65s
std
15.33s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
34.12s
p50
33.78s
p90
46.41s
p99
51.88s
min
18.33s
max
52.68s
std
10.86s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
2.73s
p50
0.00s
p90
1.57s
p99
32.09s
min
0.00s
max
36.97s
std
9.49s
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
63.5%
p50
62.0%
p90
76.2%
p99
85.8%
min
44.8%
max
86.9%
std
11.6%
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
37.0%
p50
35.0%
p90
50.6%
p99
55.4%
min
17.4%
max
55.7%
std
10.4%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.93
p50
13.00
p90
17.00
p99
17.00
min
9.00
max
17.00
std
2.91
sum
194.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).