Run started 2026-10-18 09:35 UTC
· completed 09:45 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
79.39s
p50
80.73s
p90
109.18s
p99
117.18s
min
43.35s
max
118.07s
std
24.37s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
8.99s
p50
5.68s
p90
10.20s
p99
35.38s
min
3.55s
max
38.47s
std
9.49s
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
87.82s
p50
88.99s
p90
115.66s
p99
123.66s
min
49.05s
max
124.60s
std
24.82s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
51.78s
p50
49.76s
p90
69.72s
p99
80.98s
min
31.71s
max
82.24s
std
14.06s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
34.41s
p50
25.30s
p90
61.53s
p99
68.22s
min
8.57s
max
69.01s
std
21.48s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.33s
p50
0.00s
p90
0.37s
p99
3.19s
min
0.00s
max
3.53s
std
1.02s
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
66.2%
p50
63.5%
p90
87.6%
p99
89.1%
min
50.4%
max
89.2%
std
14.1%
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
39.8%
p50
35.8%
p90
52.6%
p99
79.3%
min
15.8%
max
82.6%
std
18.2%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.00
p50
13.00
p90
15.00
p99
15.00
min
8.00
max
15.00
std
2.59
sum
144.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).