Skip to contentVeretrum Labs

Research & methods

How the model works.

An overview of the inputs, AI-assisted analysis, and reference context behind your report.

01 · Machine learning

AI vision analysis

Optional face and hand scans add image-derived geometry to your profile.

02 · Integration

Profile analysis

Your measurements and selected traits are considered together.

03 · Inference

Anatomical estimates

Length, girth, and population context in a single report.

01 AI-assisted inference

Veretrum brings body measurements, selected traits, and optional image analysis into one profile. Our proprietary estimation model considers the available information together to produce experimental estimates of phallus length and girth.

AI supports the face and hand scans. Review the detected geometry before including it, or choose an educated guess when a photo is unavailable. Image analysis and report generation take place in your browser.

The model is experimental and has not been externally validated against individual anatomical measurements.

InputProfile & images
AI visionImage analysis
ModelProfile analysis
OutputAnatomical estimates

02 Input channels

Body measurements
Height and weight provide the required starting measurements; other frame dimensions are optional. Enter exact values or use the available size estimates when unsure.
Self-reported profile
Optional descriptions add context to your physical profile. Choose the closest match and mark it as an educated guess when unsure. These descriptions do not establish current or past hormone levels.
Face proportions
Add a face photo or select the closest descriptions of your proportions. The report distinguishes reviewed photo geometry from approximate selections. Facial geometry does not establish personality or hormone status.
Digit ratio
Add a hand photo, enter a measured 2D:4D ratio, or choose a finger-length comparison. Review photo markers before confirming them. Digit ratio alone does not establish an individual’s anatomy.

03 Understanding your report

The report presents the model’s estimates from your supplied profile. Individual feature weights, interactions, and calibration details are proprietary and are not published.

Length & girth
The estimated dimensions produced for your profile, shown in inches.
Calculated volume
An approximate volume derived from the report’s predicted length and girth.
Input summary
A record of the measurements, reviewed scans, and educated guesses included in the analysis.

The result is a model output, not a measurement. Three decimal places are a display format and do not imply clinical precision.

04 Population context

See how length and circumference sit within the configured population reference. The curve provides context for the report; it does not validate an individual estimate.

Configured population reference

Explore the bell curve.

Move across the curve to explore. Drag on touch, or use ← → when the chart is focused.

+0.00σ
Percentile
50.00%

At or above: 50.00%

Length at this percentile
5.710in

14.503 cm

Circumference at this percentile
4.591in

11.660 cm

Reference density−5σ to +5σ
0.00.10.20.30.4−5−4−3−2−10+1+2+3+4+5P50.00Standard deviations from the mean (σ)
At or below the selected position Reference band · ±1σ
+0.00σ
−5σ+5σ

Length is centered at 5.71 in. Spread and circumference use the clinical reference; both measurements share the selected percentile. Extreme tails are normal-fit extrapolations.

Veale et al., 2015 · Spread reference ↗

Volume percentile in your report

Calculated volume uses the report’s predicted length and girth with a cylindrical approximation. The fixed marker and percentile place that volume within the population reference.

The volume reference is an approximation based on published length and girth distributions, assuming the dimensions are independent. It is not a published clinical volume ranking and does not account for individual variation in shape.

05 Reference context

Published studies provide population context for the report. They are not individual-level training data and do not validate the model’s predictions. Percentiles are approximate reference comparisons; extreme values are extrapolations.

  1. [1]

    Am I normal? A systematic review and construction of nomograms. (2015).

    Primary population distribution. Only 692 erect-length and 381 erect-circumference observations contributed to those endpoints.

  2. [2]

    Worldwide temporal trends in penile length. (2023).

    Contemporary sensitivity reference. Very high heterogeneity; pooled study means are not person-level training data.

  3. [3]

    Penis length and circumference according to WHO regions. (2025).

    Regional sensitivity reference. Regional and methodological differences limit individual inference.