Run started 2026-10-18 09:01 UTC
· completed 09:09 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
88.09s
p50
87.84s
p90
106.10s
p99
120.55s
min
59.88s
max
122.15s
std
20.56s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
6.30s
p50
5.35s
p90
8.55s
p99
15.26s
min
3.06s
max
16.00s
std
3.80s
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
100.06s
p50
99.73s
p90
116.63s
p99
128.49s
min
67.42s
max
129.81s
std
19.71s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
51.03s
p50
41.95s
p90
76.03s
p99
81.38s
min
34.98s
max
81.97s
std
17.47s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
42.04s
p50
38.03s
p90
58.47s
p99
70.09s
min
27.19s
max
71.38s
std
13.78s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
0.36s
p50
0.00s
p90
1.53s
p99
1.77s
min
0.00s
max
1.80s
std
0.72s
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
56.1%
p50
61.4%
p90
67.2%
p99
71.4%
min
33.3%
max
71.8%
std
12.6%
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
46.5%
p50
43.2%
p90
63.3%
p99
67.3%
min
32.9%
max
67.7%
std
11.8%
Tool Calls
Number of tool spans the agent produced for the case.
mean
11.78
p50
11.00
p90
16.00
p99
16.00
min
7.00
max
16.00
std
3.49
sum
106.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).