Run started 2026-10-18 07:46 UTC
· completed 07:52 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
88.85s
p50
89.60s
p90
118.20s
p99
140.03s
min
40.76s
max
142.45s
std
30.57s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
11.62s
p50
11.30s
p90
17.61s
p99
20.08s
min
4.74s
max
20.35s
std
4.94s
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
99.20s
p50
104.51s
p90
138.25s
p99
149.17s
min
56.60s
max
150.39s
std
32.31s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
66.22s
p50
61.64s
p90
96.28s
p99
118.11s
min
27.86s
max
120.54s
std
28.58s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
26.11s
p50
22.52s
p90
36.20s
p99
41.64s
min
16.44s
max
42.24s
std
8.80s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
2.30s
p50
2.90s
p90
4.75s
p99
5.48s
min
0.00s
max
5.56s
std
2.14s
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
68.8%
p50
67.6%
p90
80.3%
p99
84.0%
min
54.3%
max
84.4%
std
10.4%
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
30.0%
p50
29.7%
p90
42.5%
p99
44.0%
min
16.2%
max
44.2%
std
11.0%
Tool Calls
Number of tool spans the agent produced for the case.
mean
11.40
p50
11.00
p90
14.10
p99
14.91
min
8.00
max
15.00
std
2.32
sum
114.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).