Seedance 2.0 or HappyHorse 1.0: which model to pick

Frame Nest currently offers two video models. They overlap on plain text to video, and differ on sound, image input and price.

Side by side

Seedance 2.0

• Text to video only
• Audio generated with the picture
• 4 to 15 seconds, at 480p, 720p or 1080p
• Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
• 7, 14 or 34 credits per second, by quality

HappyHorse 1.0

• Text to video and image to video
• Silent
• 3 to 15 seconds, at 720p or 1080p
• Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
• 13 or 22 credits per second, by quality

A five-second 720p clip costs 70 credits on Seedance 2.0 and 65 on HappyHorse 1.0.

Pick Seedance 2.0 when

• The clip needs sound. Dialogue, ambience and effects are generated together with the picture, so there is no separate audio step.
• You want a cheap draft. 480p at 7 credits per second is the lowest price on the platform, useful for checking motion before a final render.
• You need a wide frame. It is the only model here with 21:9.

Pick HappyHorse 1.0 when

• You are starting from an image. It is the model for image to video, and the clip keeps the framing of your source image.
• You want 1080p for less. 22 credits per second against 34.
• You need a very short clip. It goes down to 3 seconds.

A simple workflow

Draft the idea at 5 seconds and 720p on either model. If it needs sound, stay on Seedance 2.0. If it is silent footage and you want it sharp, render the final at 1080p on HappyHorse 1.0.

The full specifications are on the Models page.