We test AI companion apps hands-on and grade them on one fixed method — conversation, realism, value and privacy. The number means the same thing on every page, in every country.
| App | Grade | Convo | Realism | Value | Privacy | |
|---|---|---|---|---|---|---|
| Candy.ai Research-assessed · medium | 6.7 | Review → | ||||
| Secrets.ai Research-assessed · medium | 6.4 | Review → | ||||
| Joi Research-assessed · medium | 6.3 | Review → | ||||
| OurDream Research-assessed · medium | 6.3 | Review → | ||||
| DarLink AI Research-assessed · medium | 6.2 | Review → | ||||
| Kupid AI Research-assessed · low | 6.1 | Review → | ||||
| Xotic.ai Research-assessed · low | 6.0 | Review → | ||||
| DreamGF Research-assessed · medium | 5.9 | Review → | ||||
| Dondi.ai Research-assessed · low | 5.9 | Review → | ||||
| Girlfriend GPT Research-assessed · low | 5.8 | Review → | ||||
| Lovescape Research-assessed · low | 5.8 | Review → | ||||
| Swipey Research-assessed · low | 5.8 | Review → | ||||
| Couple Me Research-assessed · low | 5.8 | Review → | ||||
| MyLovely AI Research-assessed · low | 5.0 | Review → |
Click a column to sort. Each row shows its evidence type; grades follow the published fixed method.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
AI companion known for image generation, on a subscription-plus-tokens model.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
Conversation-led AI companion with a 'Moments' media-credit system.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
Adaptive AI companion (formerly EVA AI) with multiple chat models and media credits.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
Visual-first AI companion with image and video on a coin-based model.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
Swiss-based AI companion with a cross-session memory feature.
Research-assessed · medium confidence. From public review evidence, not hands-on testing.
Popular AI girlfriend generator with strong images and low entry pricing.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
AI companion with standout voice features, at a premium price.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
EU-registered AI companion with a thin independent track record so far.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
AI companion with a free tier and a contested 'zero-log' privacy claim.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
AI companion with a very large user-generated character library.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
Freemium AI companion with chat, image and video features.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
Swipe-style AI companion noted for cross-session memory.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
Browser-based AI companion from a young UK-registered operator, on a token-metered model.
Research-assessed · low confidence. From public review evidence, not hands-on testing.
Visual-led AI companion; hit by a confirmed 2026 data breach.
What 16 apps reveal about quality, pricing and privacy — including which ones train on your chats and the one with a confirmed data breach.
Every grade comes from hands-on use against a fixed rubric. The method is public and the same for everyone.
We earn commission when you subscribe. We say so on every page. The grade is never for sale.
Apps change fast. Grades carry a last-tested date and we revisit them on a cycle.
When grades move or a new app is worth knowing about, we send one short email. Nothing else.
