I'm new to the use of AI apps. I've been trying to storyboard some WAM scripts/scenarios using a few different apps. ChatGPT has been mostly useless. It's quick to censor in many cases where there isn't any adult content. I've also used DeepSeek which seems to work much better, but it requires that I screenshot all of its' responses before it deletes them. I'm also trying to use Perplexity with mixed results. Any good suggestions? Thanks for your time and help.
If you have a computer with one of the newer video cards you might be able to run Fooocs or Flux using Forge. If you can do that then you can generate anything you want after you spend some time learning how the apps are interpreting your prompts.
It's takes some work to get up and running but it's not terribly difficult. There is a lot of helpful info on line to help get you going.
tortface2000 said: Thanks for the suggestion. I'm not up on the latest processors. Most of my work has been using iPad or other Apple equipment.
I've been using cloud compute to run local models without having a good GPU. Runpod works best for me, though I was using Google Colab before. Look for tutorials for setting up Flux (or even Wan now for video) in ComfyUI using Runpod, you'll be on the way. It's not free but it won't break the bank.
I'm really not trying to generate images or video. For now I'm just working on trying to write some scripts for wam inspired role play. Everything is just text based. Apparently much of my material is to risqué for these apps, so they censor and block responses to my requests.
tortface2000 said: I'm really not trying to generate images or video. For now I'm just working on trying to write some scripts for wam inspired role play. Everything is just text based. Apparently much of my material is to risqué for these apps, so they censor and block responses to my requests.
You'll probably find better results using custom system prompt.
Alternatively, Deepseek V3 is absolutely deranged and filthy with a custom prompt.
If you want more flexibility, I recommend installing SillyTavern as the front end and customising your system prompts and responses in there. You can generate a Gemini Key in Studio to use. Or if you prefer, use openrouter which provides free access to Gemini and Deepseek (limited) Deepseek is so cheep though, it worth paying for the small price for quicker response.
You can always use a custom instruction in ChatGPT which will help too or use the api.
Thanks for your recommendations. This sounds like the direction I'll need to take and I'll research the options you provided. At least for now I just need to be able to write down some ideas without being constantly censored and having to screenshot everything before it gets deleted.