One prompt, no timeline
Describe the subject, camera movement and light in plain language. There are no keyframes, no layers and no editing timeline to learn.
Text to video
Describe the shot, pick a frame shape, and let a verified model render it. Wan 2.2 and Seedance 1.5 Pro are available now; flagship models such as Seedance 2.5 are still completing verification. Your clip stays private in your library.

Describe the subject, camera movement and light in plain language. There are no keyframes, no layers and no editing timeline to learn.
Pick 16:9 for YouTube and websites, or 9:16 for TikTok, Reels and Shorts. Each variant is a real, tested product configuration, not a crop.
The finished MP4 is stored in your account and streamed back only to you. There is no watermark, and no public gallery link is created.
One or two sentences work best. Say what is in frame, how the camera moves, and what the light looks like.
Pick 5 seconds and 16:9 or 9:16. The verified entry model starts at 2 credits; the quote shows the exact price for your selected model.
Rendering takes roughly one to three minutes. You can close the page; the job keeps running and appears in your history.
Usually one to three minutes. The job runs on a server, so you can close the tab and come back to it in your history.
The verified entry model costs 2 credits for a 5-second 720p clip. Other models show their own price before generation. New accounts start with 3 welcome credits to try the entry model.
The credits are refunded automatically. A failed job stays in your history with a safe error code so you can retry.
Not yet. This model generates 5-second clips. Longer durations are limited by the provider model and are not offered until they are verified.