Run started 2026-10-18 07:52 UTC
· completed 07:59 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
91.23s
p50
80.50s
p90
138.61s
p99
159.37s
min
48.01s
max
161.68s
std
32.47s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
9.99s
p50
8.62s
p90
14.86s
p99
17.13s
min
3.98s
max
17.38s
std
4.37s
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
101.97s
p50
92.06s
p90
150.21s
p99
169.18s
min
56.26s
max
171.29s
std
32.64s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
65.27s
p50
61.16s
p90
103.39s
p99
123.32s
min
30.08s
max
125.53s
std
28.43s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
31.50s
p50
29.06s
p90
55.75s
p99
60.90s
min
16.58s
max
61.47s
std
14.59s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
1.82s
p50
0.00s
p90
4.83s
p99
8.36s
min
0.00s
max
8.75s
std
2.91s
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
69.1%
p50
74.8%
p90
81.0%
p99
83.1%
min
40.3%
max
83.3%
std
13.7%
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
36.7%
p50
30.4%
p90
66.1%
p99
75.1%
min
18.1%
max
76.1%
std
20.1%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.00
p50
12.00
p90
17.00
p99
18.80
min
6.00
max
19.00
std
3.74
sum
132.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).