Choosing between Swipey and its alternatives: your questions

Decide your one non-negotiable feature before you sign up for anything. If it is memory, spend the trial on a single multi-day conversation and ignore everything else. If it is imagery, generate on day one, because that is where the caps usually land hardest. Comparing apps by total generosity is meaningless — they are generous in different places on purpose. Judging them on the same axis, the one you actually care about, is the only comparison that transfers to your own use.

It depends entirely on whether you want to write or to browse. A large community catalogue means you can start talking in ten seconds and try five personalities before dinner, which suits readers who enjoy variety. It also means quality varies enormously, since anyone can publish. An app like Swipey with a thin catalogue forces you to define the character yourself, which takes ten minutes and produces something that matches what you wanted rather than something close to it.

Because the image side is genuinely behind the visual-first apps, and there is no real way to export what you build. Those two things cost it points and they would cost any app points. The score reflects the whole product, not the part it does best. On memory and conversational consistency alone it would sit higher, which is exactly why we publish a five-part score breakdown instead of one number.

Your notes and your patience, and not much else. There is no shared format for character definitions or chat history across these services, and none of them are motivated to build one. Practically, you can rewrite a persona in about ten minutes from a description you paste in yourself. What you cannot rebuild is months of accumulated context, so if you are already deep into one app, weigh that before you move for a feature you may not use.

Three evenings is enough for almost everyone. The first sitting tells you about tone and setup, which is the part every app optimises for. The second reveals whether the personality holds or starts drifting. The third exposes memory, which is the feature that actually decides whether you are still using the app in a month. Anything shorter than that and you are reviewing an onboarding flow rather than a companion.

Yes, and we recommend it — it is the only way to compare like with like. Use the same opening conversation in both, mention the same handful of specific details, and keep the sessions roughly the same length. The differences show up quickly when the input is controlled. Just avoid paying for both during the trial; the free tiers are more than sufficient to separate the winner from the runner-up.

Three things: what is stored, how long it is kept, and whether deleting your account actually removes the conversations. Then look for whether any of it is used to train models. You are handing over unusually personal material, so the answers matter more here than in most software categories. If a policy is vague on retention or on deletion, treat that vagueness as the answer and pick a competitor instead.

Paid access in this category is normally a recurring subscription managed either in the app or through whichever store handled the purchase, and it renews until you stop it. We do not publish figures because they change frequently and vary by region — read the price on the checkout screen on the day. Before your first payment, find the cancellation control and confirm it exists. If you cannot locate it in under a minute, that tells you something useful.

Every character across these services is fictional and generated, and the apps are for adults only. We do not cover any service that markets likenesses of real people, and you should treat that as a hard line rather than a preference. Beyond that, ordinary caution applies: do not share financial details, addresses or anything you would not want stored, however natural the conversation starts to feel.