Credit deduction by model
- Carat Video (LTX model): The lowest credit cost among video models
- Google Veo, Seedance, and other advanced video models: Deducts many more credits per generation
- Image generation / lip sync / upscale, etc.: Credit cost varies by model and options
Credit deduction also varies by video length and resolution.
Monthly video generation
The Carat default video model (LTX) has the lowest credit cost, so it gives you the most generations for the same credits. High-performance models like Google Veo cost far more per generation, so your available count drops sharply. Credits consumption also varies based on video length and resolution settings. You can check the credits deducted per generation for each model before requesting it: tap the sparkle icon (✨) at the bottom left of the chat box, pick a model, and the detail screen shows its per-generation cost.For current video models (Seedream, Kling, Google Veo, and so on), the maximum length per generation is 15 seconds.
Image consistency and expansion
Yes, it’s possible. Upload your desired image and request edits to expand or modify the image while maintaining the original style and consistency.Image to Video
Yes, it’s supported. We support the Image to Video feature that creates videos based on uploaded images, allowing you to keep the image’s art style or characters.Check credits per model
You can check the credits deducted per generation for each model before requesting it.- App: [Home(홈)] → Tap the [Sparkle icon(반짝이 아이콘)] (✨) at the bottom-left of the chat → Select a model → View the per-generation credits on the detail screen.
- Web: Click the [Sparkle icon(반짝이 아이콘)] (✨) at the bottom-left of the chat → Select a model → View the per-generation credits on the detail screen.
The credits shown are based on the default options. Setting a higher count, resolution, or video length than the default will deduct more credits.