toreel · FAQ
Straight answers
Questions, answered without the marketing detour
Including the ones with answers you may not like. A tool that only lists what it does well is a tool you find the limits of at the worst moment.
- Read around the subject until you have something to say
- Write each script, then rewrite it shorter
- Set up, film yourself, film it again because of the light
- Cut every clip to the words, by eye, fifteen times
- Make fifteen covers that still read at thumbnail size
- Remember to post, at the right hour, fifteen times
- Type a subject before bed
- Close the laptop
- Wake up to fifteen reels, scheduled
- Read what the last batch did, in numbers
- Type the next subject, better aimed
What does toreel actually do?
You type one subject. It researches it, writes a script for each reel, films every line with your AI avatar, cuts your own footage to the words, makes a cover for each, and leaves the batch scheduled and ready to post. Fifteen reels from one topic, unattended.
How long does a run take?
Around two hours for fifteen reels, measured end to end on a real run. Most of that is waiting on avatar generation and image models rather than on your computer — which is why runs happen overnight, and why nobody should promise you fifteen finished reels in five minutes.
Do I have to approve each step?
No. The six stages hand work to each other without asking. If a reel fails it is retried, and if it keeps failing it is dropped and the rest carry on. One bad clip does not cost the batch.
Where does the footage come from?
You upload your own. toreel cuts what you give it to the words of the script rather than pulling stock — which is the reason the output looks like your channel instead of like everyone else drawing from the same library. Clips are analysed once into usable segments and reused across the batch.
Do I have to be on camera?
Once, if you want the presenter to be you. Record a few minutes and the likeness is trained from that take; every reel afterwards is filmed without you. A single photograph works too, and so does a ready-made presenter from the catalogue — see the three routes.
What subjects does it work for?
Any. The same pipeline runs for a technical breakdown, a product review or a slow reading of an idea. There is no niche built into it — the research stage adapts to the subject rather than to a category.
What languages does it speak?
Scripts are written in English and translated into Russian, and the avatar speaks either. The interface itself is English throughout; only the content is bilingual. More languages are not built.
Can it post for me?
To Instagram Reels today. TikTok, YouTube Shorts and Facebook Reels are being connected — what that involves is worth reading if you have ever tried it yourself. Each network is linked once through an ordinary sign-in; you never handle an API key.
What does a run cost to produce?
Real money, every time. Filming fifteen avatars and generating fifteen covers are the two expensive stages, and both bill per use. Subscription pricing is not set yet — see what is known so far, which is less than you would like and more honest than a placeholder.
What can it not do?
It will not invent footage you do not have. It will not clone a person who has not consented on camera. It will not finish in minutes. And it will not make a subject interesting that was not — research makes a dull topic well-sourced, not compelling.
Is it open?
Not yet. It runs every night for one person while the rest is built, and leaving an address is currently the only way in.
Still curious
Leave an address and you will hear the day toreel opens.