I downloaded an uncensored aggressive Qwen 3.5 model and I can see in its reasoning that it is still limiting responses based on safety guardrails (e.g. violence, NSFW).
Anybody have recommendations for truly uncensored models?
EDIT: I turned off reasoning and I think it’s more uncensored if I’m very specific about what the response should include.


Jesus that MoE wiki is a fucking rabbit hole.
Thanks for sharing! Unfortunately I haven’t invested in a decent computer yet. Using 16GB GPU so been stuck on 4B Q4’s.
I’m not particularly interested in ERP, but I have obviously been using it for testing models. I’m more curious about other topics with guardrails.
I noticed that Qwen 3.5 uncensored is good if I turn off reasoning and explicitly say I want it to break the rules.
I’ll check out sillytavern tho. Thanks!