How We Review AI Story Services

What evidence do AI Porn Stories reviews use?

AI Porn Stories reviews use only evidence any visitor can see without paying: pricing pages viewed while logged out, help centres and documentation, terms and content policies, public previews, and public review profiles. We also watch every first episode we publish. We do not create accounts, pay for plans or enter personal details to gather evidence, so we describe our reviews as comparisons, never as product testing. Where a fact sits behind a sign-up, the review says it could not be verified and gives the date.

What are the 8 criteria and their weights?

The 8 criteria carry weights that add up to 100. Story video and roleplay carry the most because this publication exists for people who want stories; price and access come next because they decide whether you can actually reach those stories.

Which story criteria count most (35%)?

1. Story video and series (20%). Whether the service offers episodic or scripted video stories, how large the library is, and how clearly each episode's access is shown. An 8 needs a real series format with an access ladder visible before you pay; a 5 means occasional video without a series structure or unclear unlocks.
2. Story roleplay and scenario tools (15%). Scenario openers, steerable chats, story generators, heat or intensity controls. An 8 needs tools that shape a plot, not just a reply; a 5 is plain chat with a character.

How are access and cost criteria scored (35%)?

Access and cost criteria score what you can do before paying and what paying costs.

3. What the free tier allows (10%). How much you can judge before paying. An 8 lets you try the core experience for days; a 5 is a short sample; a 4 or lower is a handful of messages.
4. Entry price and long-term value (15%). The monthly and yearly price of the first paid plan and what it includes. Unpublished prices score low, because you cannot compare what you cannot see.
5. Clarity of token or credit costs (10%). Whether the cost of each image, voice minute, video or episode is published. A 9 means a full public price list; a 5 means the uses are listed but not the costs.

How are character, voice and policy criteria scored (30%)?

Character, voice and policy criteria cover consistency, sound and the rules a service enforces.

6. Custom characters and memory (10%). Depth of the character builder and how well long stories stay consistent.
7. Voice and audio (10%). Voice messages, calls and voiced content, and how much of it is available without extra spending.
8. Policy fit and billing discretion (10%). Whether the content policy allows the themes our readers come for, how age and real-person rules are handled, and how charges appear on a statement.

How is the overall score calculated?

The overall score is calculated as a weighted average: each criterion score is multiplied by its weight, the results are added and divided by 100, and the total is rounded to one decimal. Ranking follows the overall score. A service that cannot be scored on a criterion because the facts are hidden gets a low score on that criterion rather than an estimate, which is why unpublished prices hurt.

4 facts we could not verify on 1 November 2026

We could not verify four groups of facts on 1 November 2026, because each sits behind a sign-up:

• Token or credit costs per action on Candy AI (except its Roulette rate), Lovescape and GoLove AI.
• GoLove AI's plan prices, which appear only after registration and an age gate.
• Free-tier limits on Lovescape and the number of free episodes in GoLove AI's AI Series.
• Candy AI Premium Plus pricing and Candy token pack prices.

When any of these become public, or a service changes its plans, we update the cells, the score and the date together. The current results are on the comparison page.

What do the scores not measure?

The scores do not measure user opinion: they are not user ratings, and they are not a measure of how attractive the characters are. They describe how well each service serves someone who wants to watch, steer or write adult AI stories, judged on facts we can show.