Chris Furrey

I wanted one place where those boring details come first, written by a real person who signs his name under every page.

This page tells you plainly how I work, including the parts that are not finished yet. That matters more than any badge. If you know where a number comes from, you can decide how much to trust it. If you know what I have not done, you know where to be careful.

When an official page does not say something, I look at what other review sites report and cross-check them against each other. Those points are labelled on the page as what other reviewers report, never as my own experience.

The pages you read are built from that database. A price in a table, a plan name on a card or the line about your statement is printed from the record, not typed by hand. So when a price changes, it changes everywhere at once.

A price that has not been checked for thirty days is hidden automatically instead of being shown stale. That rule costs me a little work every month, and it is the one I care about most.

Candy AI stops the free account after three to five messages and three to five images. OurDream AI starts a new account with 55 dreamcoins. Joi AI sends its confirmation email to the spam folder and offered a new-user coupon. Each of those pages carries a dated log, and the label at the top says "Hands-on test, free tier".

A hands-on check is short and specific. It means creating a free account, reading what the app gives a new user, and looking at the logged-in pricing screens that are hidden from visitors. It does not mean paying, and it does not mean weeks of chatting. The log says exactly what was seen and when, and nothing on the page goes beyond it.

Every other app on the site is a desk review. The label says so: "Desk review - not paid for yet". I would rather tell you that than dress up a pricing page as a week of testing.

It keeps the content policy in the app's own terms, the platforms it runs on, whether you can sign up with an email only, and what the privacy policy says about training on your chats. When two sources disagree, both stay in the record with their dates until one of them is confirmed, and the page uses the careful one.

A desk review cannot tell you how warm a character feels at midnight, how well memory holds after a week, or how fast the replies come on a busy evening. When a page touches those questions, it says what other reviewers report and names that as their view.

That is also why most apps show "Not rated yet" instead of a score. A score needs time with the paid app. Until an app has had that, I would rather leave the number empty than invent one. Lists still have an order, and the affiliate disclosure explains how that order is made and where money plays a part.

You are an adult making your own choice, and you will not find a lecture here. What you will find is the price before the pitch, the catch before the button, and the free start first whenever an app has one.

I keep the tone plain and a little dry. Numbers beat adjectives: "$12.99 a month, and images cost tokens on top" tells you more than "pricey". When I give an opinion, it is my reading of the facts, and I say so. When a page is unsure, it says that too.

The verdict says who the app suits and who should skip it. Lists work the same way, only shorter, with one entry per app.

If something looks wrong, tell me. You can write through the contact page or email me at [email protected]. I fix mistakes quickly and log them on the corrections page.

The second is tokens: a monthly fee often covers chat but not images, voice or video, which cost a separate currency that runs out quickly. The third is the name on your card statement, which may or may not reveal what the charge was for. Every review puts these three near the top, and a page says "Not known yet" rather than guessing.

A partner's price, free tier, content policy and catch are printed from the same record, checked the same way, as those of an app that pays nothing. The full rules are on the editorial policy page.

The order of a list is where money can play a part, and I would rather say that than hide it. When apps fit a topic equally well, an app with a partner programme can be placed higher. What never changes is the record behind it: the same checked prices, the same free-tier numbers and the same catch, whether the app pays a commission or not.

Commercial cards never print cons because they are short sales boxes; the full picture, catches included, sits in the text and in the review.

Affiliate links carry a short tag so the partner knows the visit came from here; they do not carry your name or anything you typed on this site.

Finally the change is logged on the corrections page with the date, what was wrong and what it says now. I do not quietly edit a page and pretend it was always right. A visible correction is a small embarrassment; a hidden one is a broken promise, and I would take the first every time.

If you are the one who spotted the mistake, you are welcome to be thanked there by first name, or not at all if you prefer.

The same goes for an app that changes its rules. When a free tier shrinks or a content policy tightens, the old line is not kept for old times' sake. The record changes, the date moves, and the page tells you what the app allows now, not what it allowed when the page was first written.

I keep the answers near the top and the details lower down, so you can stop reading as soon as you have what you need.

The diary section holds my own opinion pieces about the category, written as opinions and never as test reports.

Until then, every page tells you exactly what it is: a careful reading of what each company publishes, what other reviewers report, and, for six apps, what their free tiers actually did. If that is the kind of guide you were looking for, welcome. If you were hoping for someone who has spent a fortune on every subscription, I am not that person yet, and I will not pretend to be.