Run started 2026-10-18 11:42 UTC
· completed 11:52 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
91.93s
p50
89.44s
p90
141.41s
p99
145.87s
min
24.93s
max
146.36s
std
34.85s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
11.27s
p50
6.74s
p90
16.82s
p99
33.70s
min
4.84s
max
35.58s
std
9.08s
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
102.28s
p50
100.61s
p90
154.82s
p99
157.59s
min
35.54s
max
157.90s
std
36.18s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
60.18s
p50
47.77s
p90
86.05s
p99
118.04s
min
19.44s
max
121.59s
std
27.92s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
39.11s
p50
39.29s
p90
63.56s
p99
80.64s
min
5.47s
max
82.54s
std
22.03s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.43s
p50
0.00s
p90
2.37s
p99
2.38s
min
0.00s
max
2.38s
std
0.96s
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
63.7%
p50
71.2%
p90
80.8%
p99
81.9%
min
41.3%
max
82.0%
std
14.9%
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
39.4%
p50
32.1%
p90
59.3%
p99
62.5%
min
20.1%
max
62.9%
std
16.2%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.45
p50
14.00
p90
15.00
p99
16.80
min
5.00
max
17.00
std
3.27
sum
137.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).