Run started 2026-10-18 06:51 UTC
· completed 06:56 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
83.62s
p50
68.17s
p90
136.78s
p99
154.23s
min
32.51s
max
156.17s
std
41.32s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
8.24s
p50
8.04s
p90
11.19s
p99
11.34s
min
4.89s
max
11.36s
std
2.40s
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
94.07s
p50
77.89s
p90
151.03s
p99
163.95s
min
43.30s
max
165.39s
std
42.54s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
54.33s
p50
44.36s
p90
76.94s
p99
118.20s
min
19.06s
max
122.78s
std
29.22s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
31.99s
p50
21.47s
p90
72.51s
p99
84.29s
min
14.02s
max
85.60s
std
24.39s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
1.17s
p50
0.00s
p90
2.82s
p99
5.80s
min
0.00s
max
6.13s
std
1.89s
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
63.8%
p50
59.5%
p90
77.9%
p99
85.4%
min
36.1%
max
86.2%
std
14.0%
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
36.1%
p50
38.4%
p90
52.0%
p99
60.9%
min
17.5%
max
61.9%
std
13.2%
Tool Calls
Number of tool spans the agent produced for the case.
mean
10.36
p50
11.00
p90
16.00
p99
16.00
min
5.00
max
16.00
std
3.59
sum
114.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).