Ads from product photos
Picture to video keeps the real product in every frame.
AI Based Video Generator: How the Models Work
An AI based video generator is a model trained on video that turns words, a picture or a reference clip into new footage. Reviral's Cinema Studio puts several of these models behind one prompt box, priced per clip, for example a 5-second Seedance 2.5 clip at 720p costs 115 credits (~$1.43).
No credit card required to try — sign in with email or Google.
- Text-to-video, picture-to-video and reference video
- Seedance 2.5, Wan 3.0, Veo 3.1, Kling V3 and more
- Live price per clip, before you generate
- No subscription needed
The three ways in
Text to video: you describe the scene and the model invents everything in it. Picture to video: you give a photo and the model moves it, keeping the person, product or place from the picture. Reference video: you add pictures, clips and sound the model should follow, such as a face, a dance move or a song.
Models differ in how many references they take, how long a clip can be and whether sound comes with the picture, so the model is the first choice you make.
Models on Reviral, side by side
Read from Reviral's public model list; each price cell is the live credit price for that clip length.
| Model | Clip length | References | Price for the clip |
|---|---|---|---|
| Seedance 2.5 720P | 4 to 30 s | 30 pictures, 10 videos, 10 audio | 5 s: 115 credits (~$1.43) |
| Wan 3.0 | 2 to 30 s | 10 pictures, 5 videos, 5 audio | 5 s: 99 credits (~$1.24) |
| Veo 3.1 Fast | 8 s | 3 pictures | 8 s: 24 credits (~$0.30) |
| Veo 3.1 Lite | 8 s | 3 pictures | 8 s: 12 credits (~$0.15) |
| Kling V3 Omni | 3 to 15 s | 2 pictures | 5 s: 50 credits (~$0.62) |
Read from each site's own page on 21 October 2026: Reviral public model list
How to choose
For a person who must stay the same across shots, or a clip that follows your own song, use a model with many references, such as Seedance 2.5. For the lowest price per clip, use Veo 3.1 Lite. For long single shots, Seedance 2.5 and Wan 3.0 both reach 30 seconds.
Ways to use this tool
Music-led clips
Your song as a reference so the motion follows it.
Story scenes
Text to video for places and people that do not exist.
Model tests
The same prompt on two models before you scale up.
How it works
Step 1
Open Cinema Studio
The Try button opens it with Seedance 2.5 chosen.
Step 2
Write the prompt
Subject, action, camera, light and sound.
Step 3
Add references
Pictures, clips or audio the model should follow.
Step 4
Generate
Pick the length; the price shows on the button.
Real examples
Click any example to play it. Every clip below is an existing Reviral render.
15-second Istanbul travel ad, made with Seedance 2 in Cinema Studio
Woman in a tangerine trench on a wet crosswalk, 5 s, Wan 3.0
What it costs
A 5-second Seedance 2.5 clip at 720p costs 115 credits (~$1.43).
Changing the video's length, media type, or resolution can change the final price.
Dollar figures use the $4.99 Mini Pack rate. Cached for up to 1 hour.
Frequently asked questions
- What is an AI based video generator?
- A generator that uses a video model trained on footage to create new clips from text, pictures or reference media. On Reviral a 5-second Seedance 2.5 clip at 720p costs 115 credits (~$1.43).
- Which AI video model is cheapest on Reviral?
- Among the models in the table, Veo 3.1 Lite costs 12 credits (~$0.15) for an 8-second clip, which fits inside the 100 free sign-up credits.
- Do I need a subscription?
- No. Credits are spent per clip and the price shows on the Generate button before you press it.
More in AI Video Models
Compare closely related tools, or browse the complete AI Video Models topic hub.
Lyria API: Google's Music Model, Prices and LimitsWhat Lyria is, how Google offers it through the Gemini API and Google Cloud, what each call costs, and how to put music under video.AI Human Video Generator: Real-Looking PeopleWrite the person, the movement, the camera and the light; Wan 3.0 renders a real-looking human clip.Wan Image to Video: Wan 3.0 and Wan 2.6 OnlineAdd a photo, describe the motion, render with Wan 3.0 or Wan 2.6 in the browser.Open Source AI Video Generator: Wan, Hunyuan, LTXThe main open video models, the GPU each needs, and a hosted option when you have no GPU.Chat GPT Video Generator: Prompt to Video ClipWrite the prompt in a chat, paste it here, and a video model turns it into a clip.AI Human Video GeneratorRealistic people moving, from a written scene.AI Auto Video GeneratorOne topic in, a finished narrated video out.Grok Video PromptsHow to write a video prompt that works in Grok and other models.