> For the complete documentation index, see [llms.txt](https://docs.qolaba.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.qolaba.ai/image-generation/input-controls/prompt-and-magic-prompt.md). # Prompt and Magic Prompt ### Prompt & Magic Prompt The prompt is the primary input for every image generation. It tells the model what to generate — the subject, environment, style, lighting, and mood. How you write it directly determines the quality, accuracy, and consistency of the output. *** #### Writing an Effective Prompt A strong image prompt covers four elements: | Element | What to Include | | ------------------- | ----------------------------------------------------------- | | **Subject** | Who or what is the focus of the image | | **Environment** | Where the scene takes place — setting, time of day, weather | | **Style & Medium** | Visual aesthetic, artistic approach, or rendering style | | **Mood & Lighting** | Emotional tone and how the scene is lit | **Weak prompt:** ``` A woman in a city. ``` **Strong prompt:** ``` A professional woman in a tailored navy suit walking through a rain-soaked Manhattan street at dusk, neon signs reflecting on wet pavement, cinematic lighting, shallow depth of field, photorealistic, highly detailed. ``` The second prompt gives the model a subject, environment, style, mood, and lighting — producing a far more consistent and usable output. *** #### Prompt Structure Guide Build your prompt in layers — start with the subject, then add environment, then style, then mood and quality: **1. Lead with your subject** ``` A golden retriever puppy... ``` **2. Add environment and setting** ``` ...sitting on a sunlit wooden porch surrounded by autumn leaves... ``` **3. Define style and medium** ``` ...oil painting, impressionist style, thick visible brushstrokes... ``` **4. Set mood, lighting, and quality** ``` ...warm afternoon light, soft shadows, highly detailed, 4K. ``` **Full combined prompt:** ``` A golden retriever puppy sitting on a sunlit wooden porch surrounded by autumn leaves, oil painting, impressionist style, thick visible brushstrokes, warm afternoon light, soft shadows, highly detailed, 4K. ``` *** #### What to Avoid * **Vague descriptors** — *"beautiful"*, *"amazing"*, *"cool"* add no visual direction. Replace with specific visual attributes — *"vibrant color palette"*, *"dramatic shadows"*, *"minimalist composition"* * **Contradictory instructions** — *"photorealistic watercolor"* or *"dark bright scene"* confuse the model. Pick one clear direction * **Overloading with unrelated elements** — Keep each prompt focused on one scene or subject. Complex multi-element prompts produce inconsistent results * **Describing abstract concepts without visual grounding** — *"the feeling of loneliness"* needs a concrete visual anchor: *"A solitary figure sitting on an empty park bench in winter fog"* *** #### Multi-Language Prompt Support Prompts can be written in any language — the model interprets and generates based on the language of your input. This makes Image Generation natively accessible for non-English workflows without requiring translation. *** #### Magic Prompt Magic Prompt is an AI-powered enhancement tool that takes your prompt and rewrites it into a more detailed, generation-optimized version — automatically adding descriptive language, stylistic direction, and visual detail that improves output quality. **When to Use Magic Prompt** Use Magic Prompt when: * Your prompt is short or underdeveloped * You want richer stylistic output without detailed prompt engineering * You are exploring a new subject or style and want a stronger starting point * You are not getting the visual quality you expect from your current prompt **When to Skip Magic Prompt** Skip Magic Prompt when: * Your prompt is already highly specific and detailed * You are working with reference images where strict visual consistency matters * You need precise control over exactly what the model generates * You are using a custom preset that already handles style scaffolding **How to Use It** 1. Enter your base prompt in the prompt input field 2. Enable the **Magic Prompt** toggle 3. Click **Generate** — Magic Prompt rewrites your prompt before sending it to the model 4. Review the enhanced prompt if visible — use it as a starting point for further refinement **Before Magic Prompt:** ``` A fox in a forest. ``` **After Magic Prompt:** ``` A red fox standing alertly among ancient oak trees in a misty autumn forest, dappled golden light filtering through the canopy, shallow depth of field, photorealistic, highly detailed fur texture, cinematic composition, 4K. ``` --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.qolaba.ai/image-generation/input-controls/prompt-and-magic-prompt.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.