Overview
Sora 2 is OpenAI’s video generation model, available through varg via fal. It excels at cinematic, photorealistic shots and offers a unique remix capability for transforming existing videos.| Model ID | Mode | Credits | ~Cost |
|---|---|---|---|
sora-2 | Text/Image-to-video | 100 | $1.00 |
sora-2-pro | Text/Image-to-video (premium) | 300 | $3.00 |
sora-2-remix | Video-to-video remix | 100 | $1.00 |
Quick start
Capabilities
Text-to-video
Image-to-video
Animate a still image:cURL
Video remix
Transform an existing video with a new prompt. Unique to Sora 2.Parameters
Text description of the video or transformation to apply.
Video duration in seconds.
Output aspect ratio.
For image-to-video:
[{ url: "image.jpg" }]. For remix: [{ url: "video.mp4" }].Pricing
| Model | Credits | USD | Best for |
|---|---|---|---|
sora-2 | 100 | $1.00 | General use |
sora-2-pro | 300 | $3.00 | Maximum quality |
sora-2-remix | 100 | $1.00 | Video transformation |
Tips
- Sora excels at cinematic, photorealistic content. It handles lighting, depth of field, and natural motion well.
- Use remix for style transfer, seasonal changes, or transforming the mood of existing footage.
- Pro tier is only worth it for hero shots. The standard model is usually sufficient.
- Combine with image generation — create a reference frame with Flux Pro or Nano Banana, then animate with Sora.
Related models
Seedance 2
ByteDance’s premium model with video editing and extension.
Kling V3
Best overall model. Flexible 3-15s duration.
Grok Imagine Video
xAI’s model with native audio and video editing.