Skip to content
TheChatArchiveThe Chat Archive
Amazon Novaamazon-nova-experimental-chat-05-14· AmazonHugging Face LMArena Human Preference 140K

Help me to improve this prompt for a picture generating AI: Create a photorealistic…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Claudeclaude-3-7-sonnet-20250219-thinking-32k’s answer.
  1. User

    Help me to improve this prompt for a picture generating AI: Create a photorealistic image of children age 4 to 6 in a kindergarten setting. The background is colourful and bright. Everyone is jolly happy. The children are accompained by a kindergarten teacher. She has a friendly face and smile. She also wears glasses. The children are crafting something. They are sitting around a round table, crafts are scattered on the table and on the floor. The kids use scissors and some kind of glue. Furthermore, there are craft materials with bright colours such as pipe cleaners, pompons and a stack of paper, colourful crayons. Make the scene happy.

  2. Amazon Nova Assistant

    Certainly! Here's an improved prompt for a picture-generating AI:


    Prompt:

    Create a photorealistic image depicting a lively and joyful kindergarten setting with children aged 4 to 6. The scene should exude happiness and vibrancy. Here are the detailed elements to include:

    1. Background:

      • The background should be colourful and bright, with vibrant walls decorated with children's artwork, alphabet posters, and educational charts.
      • Include colourful rugs on the floor and shelves filled with brightly coloured toys, books, and educational materials.
      • Add some indoor plants to enhance the lively atmosphere.
    2. Children:

      • There should be a diverse group of 5-7 children, aged 4 to 6, with varying ethnicities and genders to promote inclusivity.
      • Each child should have a joyous expression, some laughing, some concentrating on their crafts, and others chatting happily.
      • Their clothing should be casual and child-appropriate, with bright colours and fun patterns.
    3. Kindergarten Teacher:

      • Include a kindergarten teacher in the scene. She should be in her late 20s to early 40s.
      • The teacher has a friendly face and a warm, genuine smile.
      • She is wearing glasses that suit her friendly appearance.
      • Her attire should be professional yet approachable, perhaps wearing a colourful, comfortable top and pants or a dress with a cardigan
      • The teacher is actively engaging with the children, either helping one of them with their craft or overseeing the activity with a pleased expression.
    4. Activity:

      • The children are crafting and are sitting around a large, round table in the centre of the image.
      • The table and floor around it should be covered with various crafts in different stages of completion.
      • Visible crafting materials should include:
        • Scissors (child-safe)
        • Glue sticks or glue bottles
        • Brightly coloured pipe cleaners
        • Pompons in various colours
        • A stack of colourful paper
        • Colourful crayons and markers
      • Some children could be shown in the act of cutting, gluing, or assembling their crafts to add dynamism to the scene.
    5. Atmosphere:

      • The overall atmosphere should be happy and energetic.
      • Capture the sound of children’s laughter and chatter (visually implied through their expressions and body language)
      • Soft, natural light should pour in from large windows, enhancing the bright and cheerful environment.
      • To add more life to the scene you could include a calendar with the current month and some of the days marked with stickers indicating special activities or events

    Mood:

    • Ensure the mood of the image is light-hearted, joyful, and welcoming. Every element, from the children’s expressions to the colourful decorations, should contribute to a sense of happiness and fun.

    This prompt provides detailed and specific descriptions to help the AI generate a vivid, photorealistic image that meets all your requirements.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.