Methodology: How We Compare NSFW AI Tools

nsfw.report does not rate tools with a single number. A number hides the trade-offs that matter: one tool is cheaper, another is more private, a third remembers you better. Instead, each desk compares at least six tools on the same fixed dimensions, shows every answer side by side, and then names one overall pick.

How a comparison works

Fix the dimensions first. Each desk publishes its dimensions below before any tool is compared. Every tool is checked against all of them.
Collect the facts. We use each tool's official pricing, limits, policy and help pages, and patterns in public user reports. Every report lists its sources.
Show them side by side. Each comparison has a table with every tool and every dimension, so you can disagree with us using the same facts.
Name a pick. Every tool has real strengths. The overall pick is the tool we would recommend to most readers once all dimensions are weighed together, and the report explains why.

The dimensions, by desk

Every desk also compares starting price, free tier and platforms.

Image

Models and styles: Decides what the images can look like: photoreal, anime, illustration.
Character consistency: Whether the same person survives across a set, the hardest part of adult image work.
Editing tools: Inpainting, pose and outfit control save regenerations and credits.
Content rules: What is allowed and what is banned, in writing.
Cost per image: The real price, not the plan price.
Ownership and private mode: Who can use your images, and whether they are public by default.
Privacy and retention: How long prompts and images are kept, and whether they train models.
Age verification: How the service checks that users are adults. See who verifies what.
Setup and ease of use: Time from sign-up to a usable image.
User reports: What paying users say about billing, support and reliability.

Video

Input modes: Text, image, story or reference photos decide how much control you have.
Video length: Clip length, and whether clips join into one longer video.
Resolution and audio: Output resolution, and whether the video has sound.
Character consistency: Whether faces stay the same from shot to shot.
Content rules: What is allowed and what is banned, in writing.
Cost per 5-second clip: Video burns credits fast, so cost per second matters most.
Payment and refunds: Subscription or pay-as-you-go, payment methods and refund rules.
Privacy and retention: How long uploads and videos are kept, and whether they train models.
Age verification: How the service checks that users are adults. See who verifies what.
User reports: What paying users say about billing, support and reliability.

Voice

Product type: App, story narrator, voice chat, API or local software.
Voices and expression: Range of voices and how much emotion they can carry.
Live calls: Whether you can talk in real time, and how fast it answers.
Content rules: What is allowed and what is banned, in writing.
Voice cloning and consent: Whether real voices can be cloned, and what proof of consent is required.
Privacy and retention: How long audio is kept, and whether it trains models.
Age verification: How the service checks that users are adults. See who verifies what.
User reports: What paying users say about billing, support and reliability.

Companions

Customization: How much of the look and personality you can set.
Memory: Whether the companion still knows you next week.
Photos, voice and calls: Which media the companion can send, and live calls.
Content rules: What is allowed and what is banned, in writing.
Credits and tokens: What costs extra on top of the subscription.
Privacy and training: Whether chats train models, and how long data is kept.
Age verification: How the service checks that users are adults. See who verifies what.
Cancellation and refunds: Trial conversion, how to cancel, and refund rules.
User reports: What paying users say about billing, support and reliability.

Text

Models and context: The model and its context length decide reply quality and memory.
Characters and tools: Character library, lorebooks, scripts and group chat.
Content rules: What the filter allows and what is banned, in writing.
Message limits: Free and paid message caps, and reply speed at busy times.
Privacy and training: Whether chats train models, encryption and retention.
Age verification: How the service checks that users are adults. See who verifies what.
User reports: What paying users say about billing, support and reliability.

What we will not cover

We do not cover tools built to make sexual images of real people without consent ("undress" or face-swap tools aimed at real people), and nothing involving minors, ever. A hosted service with no stated 18+ requirement is excluded. Services that rely on self-declaration are included, and that is shown under Age verification. Local software has no account and no check, and we say so.

How often we update

We check each report again at least every 30 days. We also check it when a vendor changes its prices or its rules. The check date on every table shows when we last checked those facts.

When a fact is not public

Some vendors do not publish a price, a limit or a rule. In that case we write "Not stated", "Not published" or "Not confirmed". We never estimate prices or limits.

Corrections

If we get a fact wrong, we fix it and note the change and date at the end of the report. Send corrections to info@nsfw.report. We correct facts. We do not change verdicts on request.