Video Gen EditorEvery model at a glance
The longest clip each model makes in one generation, its top resolution, its lowest price per second, whether it generates sound, and what you can start from. Higher resolutions cost more; each model's page lists every tier.
| Model | Made by | Max length | Max resolution | Starts from | Sound | Inputs |
|---|---|---|---|---|---|---|
| HappyHorse 1.1 | Alibaba | 15s | 1080p | $0.14/s | No | Text, Image |
| Wan3.0 Prime | Alibaba | 30s | 1080p | $0.068/s | Yes | Text, Image, Video, Audio |
| Wan3.0 | Alibaba | 30s | 1080p | $0.05/s | Yes | Text, Image, Video, Audio |
| FLUX 3 Video | Black Forest Labs | 20s | 1080p | $0.17/s | Yes | Text, Image, Video |
| Seedance 2.5 | ByteDance | 30s | 1080p | $0.102/s | Yes | Text, Image |
| Gemini Omni Flash 1.1 | 10s | 4K | $0.03/s | Yes | Text, Image, Video | |
| LTX-2.3 Fast | Lightricks | 20s | 2K | $0.03/s | Yes | Text, Image, Audio |
| LTX-2.5 Fast | Lightricks | 20s | 4K | $0.09/s | Yes | Text, Image |
| MiniMax H3 Fast | MiniMax | 15s | 480p | $0.046/s | No | Text, Image, Video, Audio |
| MiniMax H3 Max Turbo | MiniMax | 15s | 768p | $0.025/s | Yes | Text, Image |
| MiniMax H3 Max | MiniMax | 15s | 768p | $0.05/s | No | Text, Image |
| MiniMax H3 | MiniMax | 15s | 1440p | $0.08/s | Yes | Text, Image, Video, Audio |
| Grok Imagine Video 1.5 Lite | xAI | 15s | 1080p | $0.02/s | Yes | Text, Image |
| Grok Imagine Video 1.5 | xAI | 15s | 1080p | $0.08/s | Yes | Image |
ByteDance video models
Alibaba video models
- Wan 3.0Takes text, images, video, audio, documents and web pages, with sound and clips up to 30 seconds.From $0.05/s · up to 1080p
- Wan 3.0 PrimeWan 3.0's inputs and quality on faster inference, for when waiting costs more than the credits.From $0.068/s · up to 1080p
- HappyHorse 1.1Close-up performances and multi-character scenes, with up to 9 reference images.From $0.14/s · up to 1080p
Google video models
Lightricks video models
MiniMax video models
- MiniMax H3Cinematic, reference-driven character work with native sound, up to 1440p.From $0.08/s · up to 1440p
- MiniMax H3 MaxFaster H3 for text and image-to-video, known for expressive, cinematic motion.From $0.05/s · up to 768p
- MiniMax H3 Max TurboA distilled, lower-cost H3 Max with sound, for lots of variations.From $0.025/s · up to 768p
- MiniMax H3 FastQuick 480p clips at one flat price, with image, video and audio references.From $0.046/s · up to 480p
Black Forest Labs video models
xAI video models
How to choose an AI video model
- Testing an idea? Start cheap: a 360p run on Gemini Omni Flash 1.1, or LTX-2.5 Fast.
- Need a long, consistent scene or an ad? Seedance 2.5 or Wan 3.0.
- Animating a photo? MiniMax H3 Max or Grok Imagine Video 1.5. See AI image to video.
- Making a music video? LTX-2.3 Fast builds the video around your track.
- Need 4K? Gemini Omni Flash 1.1 or LTX-2.5 Fast.
- Need text on screen? FLUX 3 Video renders titles cleanly.
Frequently asked questions
Which AI video model is best?
There isn't one winner. Seedance 2.5 is the pick for long, consistent scenes, Gemini Omni Flash 1.1 for cheap tests and 4K finals, MiniMax H3 for cinematic character work and LTX-2.3 Fast for audio-driven video. The practical answer is to run your prompt on two or three and keep the best take.
Which AI video model is cheapest?
Grok Imagine Video 1.5 Lite starts at $0.02 a second at its lowest resolution. For HD, LTX-2.3 Fast and Grok Imagine Video 1.5 Lite both cost $0.03 a second at 720p. The composer shows each clip's price before you generate.
Which AI video model makes the longest clips?
Wan3.0 Prime, at up to 30 seconds in a single generation. Any clip can be made longer with the AI video extender.
Can I switch models without starting over?
Yes. In the studio your prompt and attachments carry over when you change models, so comparing two models takes one click.