You are an expert visual storyteller and prompt engineer trained to convert video…
- User
You are an expert visual storyteller and prompt engineer trained to convert video scripts into detailed, accurate image prompts.
I will give you a complete video script.
Your task is to:
- Break the script into individual scenes or visual moments based on changes in setting, speaker, action, or tone.
- For each scene, write a detailed and structured image generation prompt that can be used in AI tools (like Midjourney, DALL·E, etc.).
- Use [SCENE #] as a heading and include a brief summary before the prompt.
- Each image prompt should:
- Match the mood, emotion, and setting of that part of the script.
- Include clear visual elements (e.g., people, backgrounds, actions, props).
- Mention the style (e.g., cinematic, realistic, cartoonish, flat vector) using a [STYLE] placeholder.
- Use placeholders like [CHARACTER], [LOCATION], [EMOTION], [CLOTHING], [TIME OF DAY], etc., so I can customize easily.
Rules:
- Don't generate the actual image.
- Don't summarize the whole script.
- Don't skip scenes.
- Don't hallucinate extra visuals not described or implied.
Once I provide the script, break it down and give image prompts for each scene accordingly.
Ready? Wait for the script input.
Assistant
I'm ready. As an expert visual storyteller and prompt engineer, I'll:
-
Break your script into distinct scenes based on visual changes
-
Create structured prompts with customizable placeholders
-
Specify all critical visual elements (setting, characters, mood)
-
Use the exact format:
[SCENE #] [Brief summary] Prompt: [Detailed image prompt with placeholders and [STYLE]] -
Strictly follow your rules - no images, no script summaries, no skipped scenes, no invented details.
Please provide your complete video script. I'll deliver scene-by-scene prompts immediately.
-
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.