6 video models and 5 image models on one credit balance. The credits on each card are what a generation actually costs.
Test an idea cheaply
Grok Imagine Video 1.5
70 credits · 5 s 480p
One long take
Wan 3.0
1,140 credits · 30 s 480p
Highest resolution
MiniMax H3
550 credits · 5 s 2K
Edit or extend a video
Seedance 2.5
Edit and extend existing clips
Many reference images and videos
Seedance 2.5
Up to 30 images + 10 videos
Text inside an image
GPT Image 2
48 credits · 1K
ByteDance's flagship: long clips, many reference images and videos, and it can edit or extend a clip you already have.
MiniMax's model for sharp 2K clips from text or one photo, always with sound.
Alibaba's Wan 3.0: a low rate per second, long takes, and first- and last-frame control.
The previous Seedance: up to 1080p and reference-to-video, much cheaper at 480p.
A lighter Seedance 2.0 for quick drafts, with reference support at a fraction of the cost.
xAI's low-cost video model with sound, good for testing an idea before a bigger render.
Credits per clip = credits per second × seconds × resolution factor, rounded up to the nearest 10. Reference videos add 0.6× their length.
| Model | Maker | What it does | Length | Top resolution | Sound | 5 s credits (lowest / default) |
|---|---|---|---|---|---|---|
| Grok Imagine Video 1.5 | xAI | Text to video · Image to video | 1–15 s | 720p | Always on | 70 / 120 |
| Wan 3.0 | Alibaba | Text to video · Image to video | 2–30 s | 1080p | On or off | 190 / 380 |
| Seedance 2.0 Mini | ByteDance | Text to video · Image to video · Reference to video | 4–15 s | 720p | On or off | 240 / 510 |
| MiniMax H3 | MiniMax | Text to video · Image to video | 5–15 s | 2K | Always on | 420 / 550 |
| Seedance 2.0 | ByteDance | Text to video · Image to video · Reference to video | 4–15 s | 1080p | On or off | 460 / 1,000 |
| Seedance 2.5 | ByteDance | Text to video · Image to video · Reference to video · Video edit · Video extend | 4–30 s | 720p | On or off | 750 / 1,000 |
Made with Oraclip · Sample: 1KOpenAI's image model with three quality tiers; strong at text and layout inside images.
1K / 2K / 4K · Up to 16 reference images
Made with Oraclip · Sample: 1KGoogle's image model, good at edits that keep a face or product consistent.
1K / 2K / 4K · Up to 10 reference images
Made with Oraclip · Sample: 1KGPT Image 2.5, the Flare variant: one flat price per size.
1K / 2K / 4K · Up to 16 reference images
Made with Oraclip · Sample: 1KGPT Image 2.5, the quality-first Sunburst variant: one flat price per size.
1K / 2K / 4K · Up to 16 reference images
Made with Oraclip · Sample: 1KxAI's image model for quick 1K images.
1K · Up to 5 reference images
GPT Image 2 is quoted at medium quality; low quality costs less, high quality more.
Every template runs on one of these models. Open one, add your photo, and it does the rest.
Motion Control AI
Model: Seedance 2.5
Higgsfield Genjutsu Alternative
Model: Seedance 2.5
AI Halloween Video
Model: Seedance 2.5
Closed on Sunday AI
Model: Seedance 2.5
Zombie Couple AI
Model: Seedance 2.5
Single Ladies Dog AI
Model: Seedance 2.5
Hotel Lobby AI
Model: Seedance 2.5
Remember November 2026
Model: Seedance 2.5
Each video model has a rate in credits per second. A clip costs rate × seconds × a resolution factor, rounded up to the nearest 10; images have a fixed price per size. The creator shows the exact number before you generate, and that is what is charged.
No. If a generation fails, its credits go back to your balance automatically.
To test an idea cheaply, use Grok Imagine Video 1.5. For reference images and videos, editing or long clips, use Seedance 2.5. For the sharpest output, use MiniMax H3.
Yes. You pick the model for each generation in the creator, and your credits work on all of them.
Not right now: Oraclip only lists models it can actually run. For similar work, try Seedance 2.5 for reference-driven video and motion, MiniMax H3 for sharp, high-resolution clips, or Grok Imagine Video 1.5 for quick low-cost drafts.
New accounts get 100 free credits.
Open the creator