PROMO$5 FREE CREDITS

Guides

Plain-language guides to abliteration, the Refuseless API, and the work the lineup is for.

The idea

What is abliterationAbliteration is a weight edit that reduces an LLM's refusal behavior by removing a refusal direction from the model. How it differs from a jailbreak, a fine-tune, and a system prompt.
What an abliterated LLM isAn abliterated LLM is an open-weight model whose refusal behavior has been removed from the weights. What that gives you at inference time, and what it does not guarantee.
Abliterated AIAbliterated AI is hosted inference on models whose refusal behavior was removed from the weights. What to check before you pick an API in this category.
What an uncensored LLM isUncensored LLM is a marketing phrase, not a spec. It can mean abliterated weights, a jailbreak, or a fine-tune. How to tell which one you are being sold.
An uncensored LLM APIAn uncensored LLM API is a hosted endpoint for a model that answers where stock chatbots refuse. What Refuseless actually exposes: base URL, key, model id, and retention.

Choosing

Drop in

Use cases

Glossary