{"author":{"@type":"Person","name":"Alexander Bock","url":"https://abxk.ai/en/autor/alexander-bock/"},"categories":["AI Productivity"],"content":{"markdown":"A slide needs an illustration, a newsletter needs a header image, a social post needs something that catches the eye. **Text-to-image** tools create such pictures from a written description in seconds. But the first result is rarely the one you had in mind, and not every image you can generate is one you may use.\n\nThis is part 5 of Level 1 in our AI Productivity learning path. It applies the prompt building blocks from [part 1](/en/ai-productivity/learn/prompting-basics-role-task-context-format/) to pictures. You will learn the five parts of an image prompt, a small style vocabulary and how to refine a result. You will also see where typical errors hide and what to check about rights and labels before you publish.\n\n## How an image generator works\n\nAn image generator runs on an image model. In training, the model saw a huge number of pictures with their descriptions and learned which words go with which shapes, colors and styles. When you type a prompt, it does not search for an existing picture. It generates a new one that matches your words.\n\nTwo things follow. First, the model only knows what you write. Whatever you leave out, it fills in with something typical: the subject in the center, an average style, a random background. Second, because the model learned from existing images, a result can come close to existing works, logos or characters. That is one reason why rights matter.\n\nYou do not need a special tool for your first image: ChatGPT and Google Gemini create and edit images right in the chat. Specialized generators offer more control over style and details. Our comparison lists {{\u003c stat \"area\" \"text-to-image\" \u003e}} [text-to-image tools](/en/ai-productivity/#area-text-to-image), and our [ABXK Ranking of text-to-image generators](/en/ai-productivity/best-text-to-image-generators/) shows which of them lead right now.\n\n{{\u003c tools \"chatgpt google-gemini midjourney ideogram adobe-firefly canva\" \u003e}}\n\n## The five parts of an image prompt\n\nA good image prompt answers five questions: What is in the picture? What should it look like? How is it framed? How is it lit? And which shape does it have?\n\n![A prompt for a newsletter header image, split into five labeled parts: subject, style, composition, lighting and format](/images/learn/en/create-your-first-ai-image-anatomy.webp \"Anatomy of an image prompt: five parts, five decisions you make instead of the tool.\")\n\n- **Subject:** what is in the picture, with concrete details. \"A white ceramic mug with rising steam\" works better than \"a coffee mug\".\n- **Style:** the look, such as a photo, a flat illustration or a 3D render.\n- **Composition:** close-up or wide shot, from above or at eye level, where the subject sits and where space stays empty.\n- **Lighting and mood:** soft daylight, warm evening light or studio light; calm, cheerful or dramatic.\n- **Format:** the **aspect ratio**, the relation of width to height: 16:9 is wide, 1:1 is square, 9:16 is tall.\n\nWrite the parts as plain sentences, starting with the subject. You do not need long chains of keywords.\n\nSome tools also have a field for what you do not want, called a **negative prompt** (in Midjourney: `--no text`). Otherwise, describe what should be there: \"a plain white wall\" works better than \"no posters on the wall\".\n\n## Style words that change the picture\n\nStyle is where a few words make the biggest difference. These keywords are a good starting point:\n\n| Style | Keywords to try | Works well for |\n|---|---|---|\n| Photo | \"product photo\", \"portrait photo\", \"shallow depth of field\" | Products, food, places |\n| Illustration | \"flat illustration\", \"line drawing\", \"watercolor\" | Blog headers, explainers |\n| 3D | \"3D render\", \"clay style\", \"isometric view\" | Icons, product ideas |\n| Graphic design | \"minimalist poster\", \"sticker\", \"bold shapes\" | Social posts, invitations |\n| Hand-drawn | \"pencil sketch\", \"charcoal drawing\", \"oil painting\" | Cards, mood boards |\n\nAdd one word for the light (*soft daylight, golden hour, studio light, backlight*) and one for the viewpoint (*close-up, wide shot, bird's-eye view, low angle*), and the same subject looks completely different.\n\n{{\u003c callout type=\"tip\" title=\"Describe the style, do not copy an artist\" \u003e}}\nInstead of naming a living artist, describe what you like about the look: *\"bold black outlines, flat colors, lots of white space\"*. You get more control over the result, and you do not imitate someone's personal work.\n{{\u003c /callout \u003e}}\n\n## Choose the format before you generate\n\nDecide where the image will appear before you write the prompt. A slide needs a wide image, a story on a phone a tall one. If you crop a square image to 16:9 later, you may cut off part of the subject.\n\n![Five frames with the same area but different shapes, 9:16, 4:5, 1:1, 3:2 and 16:9, each with typical uses](/images/learn/en/create-your-first-ai-image-aspect-ratios.webp \"Common aspect ratios and where they fit. All frames have the same area; only the shape differs.\")\n\nIn a chat assistant, you write the format into the prompt: \"Format 16:9\". Midjourney uses a parameter such as `--ar 16:9`, and in Canva you pick the aspect ratio before you generate, according to the vendor. If your tool does not offer the ratio you need, choose the closest one and crop with care.\n\nPlan for text, too. If a headline will go on the image, ask for \"empty space on the right\" and add the words later in a design tool. Level 3 uses the same formats for [AI video](/en/ai-productivity/learn/text-to-video-from-script-to-clip/).\n\n## Refine step by step instead of starting over\n\nYour first prompt is a starting point. Many tools return several variants at once, and in a chat assistant you can ask for another version. Pick the one closest to your idea and improve it in small steps.\n\n![Five steps: start simple, compare variants, change one thing, edit details, check and export](/images/learn/en/create-your-first-ai-image-refine.webp \"Refining an AI image: small, controlled steps instead of a new start every time.\")\n\nChange one part of the prompt per round. If you change light, style and background at once, you do not know which change helped. And when a result is almost right, do not start over: a new prompt creates a completely new image, and the good parts are gone.\n\nEdit the details instead. In ChatGPT or Gemini, you describe the change in the chat: \"Make the background lighter. Keep everything else the same.\" Many tools also let you select an area and regenerate only that part, which is called **inpainting**. Generative Fill in Photoshop, powered by Adobe's Firefly models, works this way. **Outpainting** does the opposite: it extends an image beyond its edges, for example when you need a wider format.\n\n## Common problems and where to look for them\n\nImage models have improved a lot, but some errors keep coming back. A picture that looks fine as a small preview can show strange details at full size, so always zoom in before you use it.\n\n![A sketch of a generated image with five numbered spots: hands, text, objects, faces and background, each with what to check](/images/learn/en/create-your-first-ai-image-inspect.webp \"Five spots to check before you use an AI image.\")\n\n| Problem | What you see | What helps |\n|---|---|---|\n| Hands | Six fingers, fingers that merge or bend the wrong way | Show hands doing something simple, keep them out of the frame or edit only the hand |\n| Text in the image | Misspelled words, letters that do not exist | Put the exact words in quotes and keep them short, or add the text later in a design tool |\n| Counts and positions | Four cups instead of three, left and right swapped | Ask for fewer objects, then fix the rest with an edit |\n| The same person in several images | The face changes from image to image | Use a reference image or a feature for consistent characters, if your tool has one |\n| Unwanted extras | Odd objects, fake logos, bent lines in the background | Describe the background precisely and remove extras with an edit |\n| Generic look | Smooth, interchangeable \"stock photo\" style | Add concrete details, a clear style word and a lighting word |\n\nSome tools focus on exactly these weak spots. According to the vendors, Ideogram is known for accurate text inside images, and Recraft exports editable vector graphics, which is useful for logos and icons.\n\n## Rights and labels: check before you publish\n\nGenerating an image is easy. Whether you may use it depends on the tool's terms, the rights of other people and the rules for labeling.\n\n![Three columns: uses that are usually fine, uses to check first and uses to avoid](/images/learn/en/create-your-first-ai-image-rights.webp \"A first rule set for using AI images. It does not replace legal advice.\")\n\n**The tool's terms.** They say whether you may use images commercially and who holds which rights. They differ between tools and sometimes between free and paid plans. Read them before you use an image in an ad, on a product or in a client project, and follow your company's rules on approved tools. Adobe, for example, says its own Firefly models are trained on licensed and public-domain content and calls them safe for commercial use. This does not cover the partner models that Firefly also offers.\n\n**Copyright in the result.** In many countries, including the US and Germany, copyright protects only works created by people. An image that comes from a prompt alone may have no copyright protection at all. In 2023, the US Copyright Office registered a comic book's text and arrangement, but not the Midjourney images in it (Source: US Copyright Office, decision on \"Zarya of the Dawn\", 2023, copyright.gov). This is a real risk for logos and other brand assets: talk to a designer or a lawyer before you build a brand on a generated image.\n\n**Labels.** Say when an image is AI-generated, at least when it could pass for a real photo. In the EU, the AI Act requires professional users who create a **deepfake** to disclose that the content was generated or manipulated by AI. A deepfake is an AI-made image, audio or video that resembles real people, objects, places or events and would falsely appear authentic (Source: EU AI Act, Articles 3 and 50). Many social platforms also ask you to label realistic AI content. Some tools mark their images with invisible watermarks or with Content Credentials, metadata based on the C2PA standard. Do not remove these marks. Our [AI Security](/en/ai-security/) section shows how to recognize deepfakes.\n\n{{\u003c callout type=\"warning\" title=\"Respect other people's rights\" \u003e}}\nDo not create realistic images of real people without their consent, and never upload photos of colleagues or customers without permission. Do not ask for other companies' logos or for known characters from films, games or comics. Even if the tool does not refuse, the result can violate someone's rights or damage a reputation.\n{{\u003c /callout \u003e}}\n\n## Exercise: a header image in four rounds\n\nTry the whole process with a real need, such as a header image for your next newsletter. A typical session starts with a short prompt — subject, style and format only — and ends with the full prompt from the anatomy figure:\n\n| Round | What you notice | What you change |\n|---|---|---|\n| 1 | The mug sits in the center, there is no room for a headline | Add \"mug in the left third, empty space on the right\" |\n| 2 | The light looks flat and gray | Add \"soft morning light from a window on the left\" |\n| 3 | A strange logo appears on the mug | Edit only the mug: \"plain white mug, no print\" |\n| 4 | The image works | Check it at full size, export it as 16:9, add the caption \"Image: AI-generated\" |\n\n{{\u003c callout type=\"try\" \u003e}}\nWrite a prompt for an image you really need this week. Generate it, pick the best variant and change one part per round. After four rounds, compare the first image with the last one: which part of the prompt made the biggest difference? Then check hands, text and background at full size before you use the result.\n{{\u003c /callout \u003e}}","text":"A slide needs an illustration, a newsletter needs a header image, a social post needs something that catches the eye. Text-to-image tools create such pictures from a written description in seconds. But the first result is rarely the one you had in mind, and not every image you can generate is one you may use.\nThis is part 5 of Level 1 in our AI Productivity learning path. It applies the prompt building blocks from part 1 to pictures. You will learn the five parts of an image prompt, a small style vocabulary and how to refine a result. You will also see where typical errors hide and what to check about rights and labels before you publish.\nHow an image generator works An image generator runs on an image model. In training, the model saw a huge number of pictures with their descriptions and learned which words go with which shapes, colors and styles. When you type a prompt, it does not search for an existing picture. It generates a new one that matches your words.\nTwo things follow. First, the model only knows what you write. Whatever you leave out, it fills in with something typical: the subject in the center, an average style, a random background. Second, because the model learned from existing images, a result can come close to existing works, logos or characters. That is one reason why rights matter.\nYou do not need a special tool for your first image: ChatGPT and Google Gemini create and edit images right in the chat. Specialized generators offer more control over style and details. Our comparison lists 13 text-to-image tools, and our ABXK Ranking of text-to-image generators shows which of them lead right now.\nChatGPT (GPT Image)Text-to-Image · AI at the coreOpenAI\u0026#39;s assistant; creates and edits images from a text prompt in the chat. Google Gemini (Nano Banana)Text-to-Image · AI at the coreGoogle\u0026#39;s assistant; generates and edits images with the Nano Banana models. MidjourneyText-to-Image · AI at the coreImage generator known for its artistic, stylized look. IdeogramText-to-Image · AI at the coreImage generator known for accurate text inside images, such as posters and logos. Adobe FireflyText-to-Image · AI at the coreAdobe\u0026#39;s generative AI for images and video, built into Photoshop, Premiere and other Creative Cloud apps. CanvaText-to-Image · Software with AI featuresDesign platform with a built-in AI image generator and editing tools. Short descriptions from our tool comparison, based on the vendors' information — not tested by ABXK.\nThe five parts of an image prompt A good image prompt answers five questions: What is in the picture? What should it look like? How is it framed? How is it lit? And which shape does it have?\nAnatomy of an image prompt: five parts, five decisions you make instead of the tool. Subject: what is in the picture, with concrete details. \u0026ldquo;A white ceramic mug with rising steam\u0026rdquo; works better than \u0026ldquo;a coffee mug\u0026rdquo;. Style: the look, such as a photo, a flat illustration or a 3D render. Composition: close-up or wide shot, from above or at eye level, where the subject sits and where space stays empty. Lighting and mood: soft daylight, warm evening light or studio light; calm, cheerful or dramatic. Format: the aspect ratio, the relation of width to height: 16:9 is wide, 1:1 is square, 9:16 is tall. Write the parts as plain sentences, starting with the subject. You do not need long chains of keywords.\nSome tools also have a field for what you do not want, called a negative prompt (in Midjourney: --no text). Otherwise, describe what should be there: \u0026ldquo;a plain white wall\u0026rdquo; works better than \u0026ldquo;no posters on the wall\u0026rdquo;.\nStyle words that change the picture Style is where a few words make the biggest difference. These keywords are a good starting point:\nCreate Your First AI Image: Prompts, Styles, Rights Style Keywords to try Works well for Photo \u0026ldquo;product photo\u0026rdquo;, \u0026ldquo;portrait photo\u0026rdquo;, \u0026ldquo;shallow depth of field\u0026rdquo; Products, food, places Illustration \u0026ldquo;flat illustration\u0026rdquo;, \u0026ldquo;line drawing\u0026rdquo;, \u0026ldquo;watercolor\u0026rdquo; Blog headers, explainers 3D \u0026ldquo;3D render\u0026rdquo;, \u0026ldquo;clay style\u0026rdquo;, \u0026ldquo;isometric view\u0026rdquo; Icons, product ideas Graphic design \u0026ldquo;minimalist poster\u0026rdquo;, \u0026ldquo;sticker\u0026rdquo;, \u0026ldquo;bold shapes\u0026rdquo; Social posts, invitations Hand-drawn \u0026ldquo;pencil sketch\u0026rdquo;, \u0026ldquo;charcoal drawing\u0026rdquo;, \u0026ldquo;oil painting\u0026rdquo; Cards, mood boards Add one word for the light (soft daylight, golden hour, studio light, backlight) and one for the viewpoint (close-up, wide shot, bird\u0026rsquo;s-eye view, low angle), and the same subject looks completely different.\n✦Describe the style, do not copy an artist\nInstead of naming a living artist, describe what you like about the look: \u0026ldquo;bold black outlines, flat colors, lots of white space\u0026rdquo;. You get more control over the result, and you do not imitate someone\u0026rsquo;s personal work.\nChoose the format before you generate Decide where the image will appear before you write the prompt. A slide needs a wide image, a story on a phone a tall one. If you crop a square image to 16:9 later, you may cut off part of the subject.\nCommon aspect ratios and where they fit. All frames have the same area; only the shape differs. In a chat assistant, you write the format into the prompt: \u0026ldquo;Format 16:9\u0026rdquo;. Midjourney uses a parameter such as --ar 16:9, and in Canva you pick the aspect ratio before you generate, according to the vendor. If your tool does not offer the ratio you need, choose the closest one and crop with care.\nPlan for text, too. If a headline will go on the image, ask for \u0026ldquo;empty space on the right\u0026rdquo; and add the words later in a design tool. Level 3 uses the same formats for AI video.\nRefine step by step instead of starting over Your first prompt is a starting point. Many tools return several variants at once, and in a chat assistant you can ask for another version. Pick the one closest to your idea and improve it in small steps.\nRefining an AI image: small, controlled steps instead of a new start every time. Change one part of the prompt per round. If you change light, style and background at once, you do not know which change helped. And when a result is almost right, do not start over: a new prompt creates a completely new image, and the good parts are gone.\nEdit the details instead. In ChatGPT or Gemini, you describe the change in the chat: \u0026ldquo;Make the background lighter. Keep everything else the same.\u0026rdquo; Many tools also let you select an area and regenerate only that part, which is called inpainting. Generative Fill in Photoshop, powered by Adobe\u0026rsquo;s Firefly models, works this way. Outpainting does the opposite: it extends an image beyond its edges, for example when you need a wider format.\nCommon problems and where to look for them Image models have improved a lot, but some errors keep coming back. A picture that looks fine as a small preview can show strange details at full size, so always zoom in before you use it.\nFive spots to check before you use an AI image. Create Your First AI Image: Prompts, Styles, Rights (2) Problem What you see What helps Hands Six fingers, fingers that merge or bend the wrong way Show hands doing something simple, keep them out of the frame or edit only the hand Text in the image Misspelled words, letters that do not exist Put the exact words in quotes and keep them short, or add the text later in a design tool Counts and positions Four cups instead of three, left and right swapped Ask for fewer objects, then fix the rest with an edit The same person in several images The face changes from image to image Use a reference image or a feature for consistent characters, if your tool has one Unwanted extras Odd objects, fake logos, bent lines in the background Describe the background precisely and remove extras with an edit Generic look Smooth, interchangeable \u0026ldquo;stock photo\u0026rdquo; style Add concrete details, a clear style word and a lighting word Some tools focus on exactly these weak spots. According to the vendors, Ideogram is known for accurate text inside images, and Recraft exports editable vector graphics, which is useful for logos and icons.\nRights and labels: check before you publish Generating an image is easy. Whether you may use it depends on the tool\u0026rsquo;s terms, the rights of other people and the rules for labeling.\nA first rule set for using AI images. It does not replace legal advice. The tool\u0026rsquo;s terms. They say whether you may use images commercially and who holds which rights. They differ between tools and sometimes between free and paid plans. Read them before you use an image in an ad, on a product or in a client project, and follow your company\u0026rsquo;s rules on approved tools. Adobe, for example, says its own Firefly models are trained on licensed and public-domain content and calls them safe for commercial use. This does not cover the partner models that Firefly also offers.\nCopyright in the result. In many countries, including the US and Germany, copyright protects only works created by people. An image that comes from a prompt alone may have no copyright protection at all. In 2023, the US Copyright Office registered a comic book\u0026rsquo;s text and arrangement, but not the Midjourney images in it (Source: US Copyright Office, decision on \u0026ldquo;Zarya of the Dawn\u0026rdquo;, 2023, copyright.gov). This is a real risk for logos and other brand assets: talk to a designer or a lawyer before you build a brand on a generated image.\nLabels. Say when an image is AI-generated, at least when it could pass for a real photo. In the EU, the AI Act requires professional users who create a deepfake to disclose that the content was generated or manipulated by AI. A deepfake is an AI-made image, audio or video that resembles real people, objects, places or events and would falsely appear authentic (Source: EU AI Act, Articles 3 and 50). Many social platforms also ask you to label realistic AI content. Some tools mark their images with invisible watermarks or with Content Credentials, metadata based on the C2PA standard. Do not remove these marks. Our AI Security section shows how to recognize deepfakes.\n!Respect other people\u0026#39;s rights\nDo not create realistic images of real people without their consent, and never upload photos of colleagues or customers without permission. Do not ask for other companies\u0026rsquo; logos or for known characters from films, games or comics. Even if the tool does not refuse, the result can violate someone\u0026rsquo;s rights or damage a reputation.\nExercise: a header image in four rounds Try the whole process with a real need, such as a header image for your next newsletter. A typical session starts with a short prompt — subject, style and format only — and ends with the full prompt from the anatomy figure:\nCreate Your First AI Image: Prompts, Styles, Rights (3) Round What you notice What you change 1 The mug sits in the center, there is no room for a headline Add \u0026ldquo;mug in the left third, empty space on the right\u0026rdquo; 2 The light looks flat and gray Add \u0026ldquo;soft morning light from a window on the left\u0026rdquo; 3 A strange logo appears on the mug Edit only the mug: \u0026ldquo;plain white mug, no print\u0026rdquo; 4 The image works Check it at full size, export it as 16:9, add the caption \u0026ldquo;Image: AI-generated\u0026rdquo; ▸Try it\nWrite a prompt for an image you really need this week. Generate it, pick the best variant and change one part per round. After four rounds, compare the first image with the last one: which part of the prompt made the biggest difference? Then check hands, text and background at full size before you use the result.\n","words":1960},"date":"2026-10-14","description":"How to describe subject, style, composition, light and format in an image prompt, refine the result, fix typical errors and check rights before you publish.","image":"https://abxk.ai/images/cover/create-your-first-ai-image.webp","json":"https://abxk.ai/en/ai-productivity/learn/create-your-first-ai-image/index.json","kind":"page","language":"en","learning_path":{"category":"ai-productivity","level":1,"level_name":"Beginner"},"modified":"2026-10-14","related_tools":["ChatGPT (GPT Image)","Google Gemini (Nano Banana)","Midjourney","Ideogram","Adobe Firefly","Canva"],"section":"ai-productivity","tags":["text-to-image","image-prompts","ai-images","beginner"],"title":"Create Your First AI Image: Prompts, Styles, Rights","translations":[{"json":"https://abxk.ai/de/ai-productivity/learn/create-your-first-ai-image/index.json","language":"de","url":"https://abxk.ai/de/ai-productivity/learn/create-your-first-ai-image/"}],"url":"https://abxk.ai/en/ai-productivity/learn/create-your-first-ai-image/"}