Text to video, made for the feed
The clip you can picture is one sentence away. Describe who is in frame, where they are, what the camera does and what they say, and Text to Video renders a short clip with sound — vertical for Reels, Shorts and TikTok, wide for YouTube. Four to fifteen seconds is the length of a hook, a transition or a b-roll insert, and that is exactly what comes back.
The page below opens the Studio with a starter prompt and 9:16 selected. Draft is the tier for testing motion at 1 credit per second; Standard is the one you post; Pro is for the hero clip at 1080p. The total shows on the button before you render, and Standard or Pro clips that get blocked or fail are refunded.
Starter prompt
“Handheld selfie-style clip of a woman with a blonde bob walking through a rainy night market, neon signs, she says "this is the best street food in the city"”
What this solves
- Hooks and transitions you can picture but cannot film today
- B-roll under a voiceover that stock libraries never quite have
- Vertical clips for three platforms from one sentence
- Spoken lines without a mic, a room or a second take
The steps
Write one sentence: subject, place, camera, line in quotes
Specific prompts give specific clips — name the light, the movement and the mood.
Text to VideoRender on Draft, then re-run the keeper on Standard
Draft costs 1 credit per second, so testing three ideas is cheaper than one Standard miss.
Text to VideoAdd a track from Music & Soundtracks
A 30-second bed for 2 credits finishes the clip without a licensing search.
Music & SoundtracksQuestions
How specific should the prompt be?
Name the subject, the setting, one camera move and the mood; put spoken lines in quotes. Fifteen to forty words is the sweet spot — long paragraphs dilute the motion.
Which resolution for TikTok?
Standard at 480p is the cost-effective choice; 720p if the clip is mostly faces. TikTok compresses uploads anyway.
Can I keep a character consistent across clips?
Yes — pick up to nine reference images from your creations, or build an AI influencer and use AI Influencer Videos for a locked face.
Text to video, made for the feed
Open with these settings