AI video models
The video models we can run are organized into series. All available with one subscription.
8 series
Seedance
ByteDanceByteDance’s video family: from quick draft to finished 1080p film with native audio.
- Seedance 1.5 Pro
- Seedance 2.0
- Seedance 2.0 Fast
Kling
KuaishouKuaishou's film family is famous for its physical dynamics that still hold up even under long lenses.
- Kling 2.6
- Kling V3 Turbo
- Kling 2.5 Turbo Pro
- Kling 2.6 Motion Control
Veo
Google DeepMindThe flagship of Google DeepMind. Dialogue and sound effects are generated by the model itself, not added after production.
- Veo 3.1
- Veo 3.1 Fast
Sora
OpenAIOpenAI's video model natively generates information and is highly responsive to prompts.
- Sora 2
Tongyi Wanxiang WAN
AlibabaAlibaba’s open source weighted film family. Cost-effective per second, 2.6 with native audio.
- WAN 2.6
- WAN 2.6 Flash
Hailuo
MiniMaxThe MiniMax family specializes in stylized dynamics and character performances.
- Hailuo 02 Pro
- Hailuo 2.3
PixVerse
PixVerseIt's fast, cheap, and has an extension mode for extending existing clips.
- PixVerse V6
- PixVerse V6 Extend
Topaz
Topazimages and videos are enlarged and charged according to the output size.
- Topaz Video Upscale