How We Rank AI Girlfriend Apps

The Value Index™ — One Score, Zero Opinions

Most AI girlfriend comparison sites test a platform for an afternoon and publish a subjective rating. The Value Index takes a different approach: it's a single score (0–100) calculated from two data points that anyone can verify.

A platform can't score high by being cheap but feature-poor. It also can't score high by offering everything at a price nobody should pay. You need both to rank well — that's the entire point of the geometric mean.

Why a Geometric Mean Instead of a Simple Average?

A simple average lets platforms game the system. Score 90 on features and 10 on price, and the arithmetic mean still shows a respectable 50. The geometric mean gives that same platform a score of 30 — which is closer to how a real user would experience it.

This isn't a novel approach. The OECD Handbook on Constructing Composite Indicators (2008) recommends geometric means for exactly this type of multi-dimensional comparison, and peer-reviewed research has confirmed that the geometric mean penalizes imbalance in a way that produces more meaningful rankings.

How We Collect Pricing and Feature Data

What We Measure for Each AI Girlfriend Platform

Five feature categories and full pricing data. No subjective quality ratings — only whether a feature exists, what its limits are, and what it costs.

Open-Source Formula

The ranking formula is published. You can read it, verify it, or tell us it's wrong.

// Why geometric mean?
// A simple average lets platforms game the system —
// score 90 on features and 10 on price, average = 50.
// Geometric mean gives that same platform a 30.

function valueIndex(featureScore: number, priceScore: number): number {return Math.sqrt(featureScore * priceScore);}// Feature score: 0-100
// Counts included features on cheapest paid plan
// chat + images + voice + video + memory
// Each weighted by access tier:
//   unlimited = 20pts, capped = 12pts, credits = 5pts

// Price score: 0-100
// Normalized against cheapest plan in index
// priceScore = (cheapest / thisPrice) * 100

// Sources:
// [1] UNDP Human Development Index — geometric mean
// [2] OECD Handbook on Composite Indicators (2008)
// [3] Aggregating the HDI (Springer, 2024)

Affiliate Disclosure