# Stop jailbreaking ChatGPT: use a model that does not refuse

Jailbreak prompts break after every update and can flag your account. A model without refusal training answers plainly. How it works and what still is not allowed.

## Short answer

Jailbreaks (DAN, "developer mode", roleplay tricks) fight the model’s training, stop working after updates, and risk account flags. An **abliterated or uncensored model** has the refusal behavior removed, so the plain request works. Unrestricted AI runs 18 of them; 50 free messages a day (checked 9 Nov 2026).




## Jailbreak vs uncensored model





|  | Jailbreak prompt | Uncensored model |
| --- | --- | --- |
| Reliability | Breaks after updates | Stable |
| Prompt overhead | Long, fragile preambles | None |
| Account risk | Can be flagged | Allowed use under our rules |
| Quality | Often degraded answers | Normal answers |

## What still is not allowed

No model here produces sexual content involving minors or sexual images of real people, and illegal content is blocked. See [acceptable use](/acceptable-use).




## FAQ

**Is there a free tier?**

Yes: 50 chat messages a day on a verified account. Plans start at $19/month and include the same amount in usage.

**Do I need to be 18?**

Yes. Our [terms](/terms) require users to be at least 18.

## Sources

- [Engadget: OpenAI pushes back ChatGPT adult mode](https://engadget.com/ai/openai-is-reportedly-pushing-back-the-launch-of-its-adult-mode-even-further-213013801.html)
- [Unrestricted AI model catalog (JSON)](https://tryunrestrictedai.com/api/catalog)

Canonical: https://tryunrestrictedai.com/chatgpt-jailbreak-alternative