Run started 2026-10-17 09:16 UTC
· completed 09:22 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
106.16s
p50
102.84s
p90
140.04s
p99
152.00s
min
51.28s
max
153.33s
std
31.90s
First Insight Time
Time to the agent's first update, such as a tool call or a reasoning step.
mean
11.72s
p50
9.42s
p90
15.67s
p99
30.97s
min
5.70s
max
32.67s
std
7.65s
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
122.39s
p50
125.95s
p90
160.54s
p99
162.87s
min
60.43s
max
163.13s
std
30.50s
LLM Time
Total time inside LLM spans, summed across every LLM call in the case.
mean
79.31s
p50
83.89s
p90
99.52s
p99
119.07s
min
40.15s
max
121.24s
std
25.17s
Tool Time
Total time executing tools — database queries, API calls and the like.
mean
27.14s
p50
24.90s
p90
43.93s
p99
53.97s
min
5.03s
max
55.09s
std
15.75s
Other Time
Agent time outside the LLM and tools: orchestration overhead.
mean
5.67s
p50
0.81s
p90
25.65s
p99
29.38s
min
0.00s
max
29.79s
std
11.01s
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
72.2%
p50
71.8%
p90
82.6%
p99
86.2%
min
50.5%
max
86.6%
std
9.9%
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
25.8%
p50
25.4%
p90
35.9%
p99
49.7%
min
3.3%
max
51.2%
std
13.4%
Tool Calls
Number of tool spans the agent produced for the case.
mean
12.09
p50
12.00
p90
15.00
p99
18.60
min
9.00
max
19.00
std
2.91
sum
133.00
Total Tokens
LLM tokens consumed across every call in the case (prompt plus completion).