Text to video
Text to video for ads: coming soon
Text-to-video turns a written description into a short clip. Adlark will offer it with Veo 3.1 Fast and Quality. It is not live yet. Here is how text-to-video prompts work for ads and what to build while you wait.
Status
Text-to-video is coming soon and cannot be used yet. When it is switched on, you will write a prompt, pick a length (4, 6 or 8 seconds) and an aspect ratio, and get a clip with sound. You will not be billed for video until then.
When text-to-video is the right choice
Text-to-video is best for shots where the exact product does not need to appear, or appears small: mood shots, B-roll, environments, a person reacting, abstract openers. For shots where your product has to look exactly right, start from an image instead (see image to video).
- Opening hooks: a strange or striking scene that earns the first two seconds.
- Lifestyle B-roll: a kitchen in the morning, a gym at night, a desk with a coffee.
- Problem scenes: tangled cables, a messy bathroom shelf, a tired face at a screen.
- Transitions between product shots.
How to write a video prompt
Good video prompts read like a single shot description from a director. Cover these five things, in this order:
- Subject: who or what is in frame.
- Action: the one thing that happens.
- Setting: where, and what time of day.
- Camera: framing and movement (static, slow dolly in, handheld, top-down).
- Style and sound: lighting, color, and the audio you expect (room tone, a pour, quiet music).
A tired office worker rubs her eyes at a laptop late at night, blue screen light on her face, dark apartment, static medium close-up, quiet hum of a fridge.Then test the same idea as a still first. Paste the prompt into the image generator with FLUX.1 schnell at 4 credits an image. If the still does not land, rewrite the prompt before spending video credits.
What to build now
| You need | Use today | Cost |
|---|---|---|
| The words | Ad copy or a UGC script | 5 credits per set, 5 per script |
| The shot plan | Script Runner | 5 credits per script plus each scene image |
| The voice | Voiceover | 13 credits per 1,000 characters |
| Captions on footage | AI Captions | 25 credits per video |
Honest limits of current text-to-video
Today's models handle short single shots well. They still struggle with readable text, exact brand logos, consistent faces across separate clips, and complex physical actions. Plan ads as several short shots with captions and voiceover carrying the message, rather than one long clip doing everything. Our guide AI video ads: what to expect goes into this.
Questions
+Can I generate video from text in Adlark today?
No. Text-to-video is coming soon. You can write scripts, storyboard them, record voiceovers and caption existing footage now.
+Which models will power it?
Veo 3.1 Fast (32 credits per second) and Veo 3.1 Quality (84 credits per second), via the Gemini API. Prices may change before launch.
+Will clips include sound?
The planned Veo 3.1 models generate audio with the clip. You can also add your own voiceover.
Related
Make your next ad today
Pro is $39.99 a month with 3,000 credits. Cancel anytime.