cardex

AnonymousDegenerate

Chub· 4 cards· 0 lorebooks· active –

I'm nobody important, just here to spread the word of transformation, suits, ABDL, and more with my bots, on this site filled with deplorable stuff. Fair warning, my fetishes are all over the place. So don't go complaining to me if you see something that you don't like, remember that you can block any tags, right?

No schedule, I release bots whenever the mood strikes, and said bots reaches a acceptable state to be released. I don't do requests, as I like to do my own thing.

Remember that everything I create is a complete work of fiction, and does not reflect what I believe, nor I condone. I shouldn't have to be saying this in the first place, but here we are.

Also remember to always support human artists and writers, human creativity will always be better than whatever AI can churn out. After all, their very works were fed to the AIs! I just like making chatbots.

You can also reupload my chatbots onto other sites if you want to. Just don't try to claim them as your own (as that's a dick move), and be sure to link back to the original if possible.


MODEL REVIEW: DeepSeek V3 0324 - Best LLM I've used so far, follow instructions easily, and can handle complex scenarios where other models would struggle. Although one has to fight against it with placing asterisks on everything, also does the "somewhere, a person does something something.". DeepSeek V3.1 - Similar to V3 0324, but has a tenacity to produce shorter responses, prone to having similar rerolls, and sometimes ignore instructions, so it has to be nudged in the right direction. At least you don't have to fight it placing asterisks everywhere anymore, as it does pick up on formatting easily, and has dropped the whole "somewhere" thing. DeepSeek V3.2 - The successor to V3.1, improved in some ways and gained some new issues. Now prone to parroting what the user just said, varies to presets and chatbots, from rarely happening, to happening too much to be bearable. Can increasingly get wordy, bad for users trying to converse tokens with keeping responses limited. It's wayyyy cheaper than anything else on the market. Great if you want the most bang for your buck. Also if you pushed the model to its absolute limit with the weirdest shit, expect it to break. Not so great if you got it stuck parroting.

4 cards · page 1 of 1