"Uncensored" is a marketing word, not a specification. Every platform in this category filters something, and nearly all of them advertise the opposite. The useful question is not whether a filter exists but where it sits, whether it applies consistently, and what happens to your credits when you hit it.
The three lines every platform draws
| Line | Universality | Note |
|---|---|---|
| Minors | Universal and absolute | Non-negotiable everywhere, and correctly so |
| Real identifiable people | Common but inconsistent | Tied to real legal exposure |
| Assorted platform-specific limits | Highly variable | Where "uncensored" claims actually break |
Why the filter is inconsistent
Most platforms run a base model plus a moderation layer, and the two disagree. The model will happily produce something the moderation layer then blocks, which is why refusals feel arbitrary and why the same prompt can pass on Tuesday and fail on Thursday. Model versions also change without notice, so the line moves under you.
The part that costs money
A refusal mid-generation may still consume the credit. This is rarely documented and is worth testing before you buy a large pack. See how credit pricing works and how to test before paying.
Part of the full NSFW AI safety checklist.
FAQ
Is any platform genuinely unfiltered?
No. The minor line is universal and absolute. Any platform claiming zero filtering is either lying or one you should not use.
Why did it refuse something it allowed yesterday?
Model version changes and moderation-layer updates, neither of which is usually announced. Inconsistency is the norm, not a bug you can report.
Do refusals cost credits?
Sometimes, and it is rarely documented. Test with a small pack before committing to a large one.