Thanks for getting back to me. However, it seems I might be using a different interface because, in my version, there are no separate fields for 'scenario' or 'situation'. I only have a single field for the character description, which is capped at 150 characters initially, and then 300 characters for image generation (which, for some reason, is separate from the description). This might also explain why I can't generate 2 or 4 images at once, and why I'm forced to generate an image every single time before I can generate a video.

One star more than Candy.ai, but character generation still sucks hard: impossible to import any reference images, and you're capped at just 150 characters for the entire description (character + scenario + situation). On top of that, the image generation has insane censorship — even harmless words like 'girl' get flagged. And don't get me started on the roleplay glitches... the AI will say 'I'm here!' describing the arrival scene in detail, then immediately follow up with 'I'll be right over' like it forgot it just showed up. Super immersion-breaking.

Interesting idea, not really ready for a full exp. Images are still with limbs/fingers problems. Conversations are not to o1/sonnet lvl.