iii Harness benchmarks
Release intelligence

Quality you can trace,
release by release.

A durable view of agent quality, reliability, cost, and scenario runtime across every relevant Harness deployment.

Last benchmark sync
Historical data connected
0 releases No subject
Latest evaluated release

Pending
Release details
Latest release matrix

Scenario metrics

Compared with the prior release — coverage
Scenario Status Quality Efficiency Reliability Details
Median Target Pass rate Δ score Cost Δ cost Runtime Hard gates Technical Missing Retries
Release trend

Quality score

Audit trail

Release history

100% report coverage
Release Status Pass rate Quality score Cost Duration Commit