You choose the prompt, the input files and the model. You are responsible for having the right to use everything you send and to use the result the way you intend. We do not check in advance whether you have that right.
It is forbidden to generate in the Studio, or to send into it: the likeness, voice or body of a real person without that person's consent, public figures included; any sexualised or abusive content involving minors; content that incites violence, hatred or a criminal offence; someone else's copyrighted work or protected mark that you have no right to use; content that breaches the terms of the model providers.
The model providers are fal.ai, Google and BytePlus, and through fal.ai also ElevenLabs. Your prompt and your input files are forwarded to them so that the generation can exist at all, so their terms of use and acceptable-use policies apply to your content just as these do. Breaking their rules breaks these terms.
The names and marks of the models and their makers belong to their respective owners and are shown in the Studio solely to identify the models; this implies no affiliation, sponsorship or partnership with those makers.
Every prompt passes an automatic check before it is sent and may be rejected. That check is a coarse first line, not a permission: a rejected prompt certainly did not pass, but an accepted prompt does not mean the content is allowed.