[ PERFORMANCE DATA ]

Fast. Accurate.
Measurably better.

Every number below comes from real scans across our beta cohort. No synthetic benchmarks, no cherry-picked repos — median figures across 100+ teams.

Updated April 2025 · Beta cohort data
< 3 minFull repo scanavg across all repo sizes
95%Dependency accuracyvs manual audit baseline
44%Fewer API callsvs comparable tools
2.8 minPlain-language outputper module, from raw scan

Scan Speed by Repo Size

< 10k lines0.4 min
10k – 50k1.2 min
50k – 200k2.1 min
200k – 500k2.8 min
> 500k lines< 5 min

Median scan duration · beta cohort · April 2025.

Dependency Accuracy vs. Alternatives

% accuracy
95%
Anchoria
71%
Tool A
63%
Tool B
58%
Manual

Accuracy measured against manual audit on 30 open-source repos.

[ METHODOLOGY ]

How we measure

Step 01Real repos only

All benchmarks use public open-source repos (React, Django, Express, Kubernetes) and anonymised private beta repos. No synthetic or toy codebases.

Step 02Ground-truth comparison

Dependency accuracy is measured by comparing Anchoria output against a hand-audited ground truth produced by two senior engineers independently.

Step 03Median, not mean

We report median figures. Outlier repos (monorepos > 2M lines) are flagged separately. The p50 number is what most teams will experience.

Want to run Anchoria on your own repo and see the numbers yourself?

Start free →