Creative Cases

These examples move from simple to practical and help you build node-based intuition.

Case 1: One-sentence image generation

Goal: quickly create a style frame.

1. Build Input Text -> Prompt -> Result.
2. In Prompt, choose image modality.
3. Pick a text-to-image capability (for example Banana / Nano Banana 2 / Seedream).
4. Enter prompt and run.

Example prompt:

A cyberpunk street market at night, neon reflections on wet pavement, cinematic, highly detailed

Case 2: Turn an image into video

Goal: animate a still image.

1. Add an Input Image node and upload an image.
2. Create a Prompt node and connect the image input.
3. Choose a video capability (for example image-to-video) and set duration.
4. Run and review output in Result node.

Model examples: Veo, Hailuo, Kling, Seedance, Wan (actual options depend on your selector).

Case 3: Analyze first, generate second

Goal: improve prompt quality via analysis.

1. Upload a reference image.
2. Use a text capability (Gemini or Seed 2.0 Pro) to analyze and describe it.
3. Feed that output into an image/video Prompt node.
4. Generate with clearer guidance and better consistency.

Case 4: Batch outputs and compare

Goal: produce multiple variants in one run.

1. Set Prompt batch size to x4.
2. Keep key parameters consistent (ratio, duration, resolution).
3. Run and compare grouped outputs.
4. Keep the best variant and continue downstream (upscale/enhance/voice, etc.).