Why show the two prompts separately?
Because the image prompt and the motion prompt solve different problems. Keeping them separate makes it easier to copy the right text into the right model.
GPT Image 2 + Seedance 2 / Prompt Gallery
Browse paired GPT Image 2 and Seedance 2 prompts with reference images, playable videos, copy-ready text, and workflow notes for image-to-video creation.
More prompts load as you scroll
How to pair GPT Image 2 and Seedance 2 prompts
These cards pair one image-generation prompt with one motion prompt so you can study how a still visual setup becomes a playable clip.
Use the GPT Image 2 section to understand composition, materials, and subject anchors, then use the Seedance 2 section to study what motion, timing, and camera instructions were added on top.
How to write stronger paired prompts
- Keep the core subject, mood, and framing stable between the image prompt and the video prompt.
- Let the GPT Image 2 prompt carry most of the visual specificity, then let the Seedance 2 prompt focus on motion, pacing, and camera behavior.
- When iterating, change one layer at a time so you can tell whether the improvement came from the still image setup or the motion instruction.
Prompt FAQ
Should the video prompt repeat the full image description?
Usually no. Keep the visual anchors that must remain stable, then spend the rest of the video prompt budget on motion, timing, and camera changes.
What should I do when a clip drifts away from the reference image?
Tighten the subject anchors in the image prompt first, then simplify the motion prompt so it asks for fewer simultaneous changes.