The questions people actually ask before joining a waitlist, answered without hedging: when access opens and how the waitlist works, what it will cost, who owns the films you make with it, how long a film can be, whether you need filmmaking experience, and which models sit underneath.
Y-O-X is not open yet. That is stated plainly here rather than buried three scrolls down, because the most common and most justified complaint about AI film tools is that availability and pricing stay deliberately vague until after you have handed over an email address. Knowing that something is not ready is useful information, and withholding it does not make the waitlist more valuable.
Where an answer is not final, it says so and explains what it depends on. Where it is final — ownership, the zero training policy — it is unambiguous, and the same commitment appears in the terms rather than only in the marketing copy. Anything on this page that later turns out to be wrong will be corrected here rather than quietly removed, which is the only version of a FAQ that is worth reading.
Access opens in November 2026. Invitations go out in batches, in order of signup date - so joining the waitlist earlier means an earlier invitation.
Plans are indicative and subject to change. Free at $0/month covers starter templates, 3 projects, 50 MB storage and 5,000 credits. Pro at $19/month adds the full template library, unlimited projects, 1 GB storage and 20,000 credits. Max at $99/month gives 10 GB storage and 125,000 credits. Enterprise is custom, with on-premise deployment, SSO and SAML, custom integrations and SLAs. Credits can also be bought outright and roll over for 12 months - one wallet covers text, audio, image and video.
You do. You own the output in full and can distribute or sell it commercially. Your scripts, prompts and generated media never enter any foundation training pipeline.
Short-form today. El Comandante runs 2:13 and the system is proven at that length - not twenty minutes, not ninety. Ask again in six months.
No. You direct it in plain language, typed or spoken - there is no timeline to master and no panels to arrange. Anyone who can describe what they want should be able to use it.
Deliberately unstated, because it changes. Y-O-X is not married to any provider: it uses the best available model for each job - picture, voice, music - and swaps them as they are beaten. That happens under the hood and should be irrelevant to you. The models keep improving; the way you work does not change.
No. Y-O-X has a zero training policy: your scripts, your direction and your finished films are never used to train any model, ours or anyone else's, and they are not passed to model providers for training either. You keep the copyright in everything you make, with no revenue share and no licence back to us, and each production sits in its own encrypted workspace so one cannot see another. That is the answer most AI tools will not put in writing, and it is the one that decides whether a tool can be used on work under NDA or headed for a festival.
Both refer to the same company and the same product. The name is Y-O-X; written without the hyphens it is YOX. The command line tool is yox, which is where you will see it unhyphenated most often.