Overview

Last 24 hours

Throughput, latency, quality, and spend across Acme Agents.

Live (auto-refreshing in prod)
Spans
1.1k
240 traces
Error rate
3.6%
p50 latency
1214ms
across all spans
p95 latency
4712ms
tail latency
Hallucination
0.81
41 critical · 73 warning
Cost
$18.03
952.0k tokens

Spans per hour

Last 24 hours · errors stacked in red

018355370-23h-17h-11h-5h

Quality at a glance

481 evaluations

Mean score
0.812/ 1.000 (RAGAS faithfulness)
Critical
41
8.5% < 0.5
Warning
73
15.2% < 0.7
Open Quality dashboard →

Top operations

By span volume

OperationCountError ratep50p95
anthropic.messages.create3311.5%1504ms2370ms
openai.chat.completions.create1500.7%1520ms2364ms
agent.run1375.8%3309ms6189ms
tool.calculator911.1%334ms619ms
tool.code_exec892.2%349ms628ms

Recent traces

Newest first

TraceStatusDurationTenantStarted
agent.run
trace_00000l
ok1213msglobexjust now
agent.run
trace_00005h
ok2518msacme10m ago
workflow.execute
trace_000040
ok2582msglobex29m ago
agent.run
trace_00004h
ok5196msglobex32m ago
agent.run
trace_00005m
error4152mssoylent39m ago
agent.run
trace_000055
ok3398mssoylent45m ago
agent.run
trace_00006e
ok1493msacme49m ago
workflow.execute
trace_000038
ok1933msacme53m ago