AI IMAGE TOOLS

What is Midjourney?

Midjourney is (maybe the most) a popular AI image generation tool known for its distinctive, artistic, and often surreal visual style. It transforms text prompts into high-quality visuals, making it a favorite among designers, storytellers, and concept artists.

Originally, Midjourney was accessible only through Discord, where users entered prompts including "prompt code" into a shared server and received images back from the bot. This format encouraged community interaction and prompt sharing but also was less user friendly.

As of 2024, Midjourney is also available via a dedicated web platform, which offers a more private and focused workspace for image creation. The web version includes easier access to your prompt history, galleries, and settings—ideal for solo projects or professional use.

• Discord Version: Community-driven, prompt collaboration, great for experimenting and getting feedback, downside - knowledge of and understanding "prompt code".
   
• Web Version: Cleaner interface, private generations, better suited for individual workflows and project management and much easier to use.
   

Regardless of platform, Midjourney continues to emphasize stylized visuals and excels at producing rich textures, abstract compositions, and concept-driven artwork. It’s less about realism and more about mood, imagination, and visual storytelling.

As of 2025 - Midjouney can how also transform your generated images to video. 

• Subject/Theme: Clearly describe your main subject.
• Details: Specify colors, textures, and mood.
• Art Style: Include any artistic styles or influences.
• Composition and Camera: Add perspective, camera angle, or shot type.
• Parameters: Apply technical codes (e.g., resolution or aspect ratio) or set this in your settings if you are using the browser version of Midjourney.

Example Prompts:

1. "A peaceful sunset over a mountain lake, warm colors, realistic detail, high resolution (adding "--ar 16:9" if using discord and "--r 4". On desktop version set these under your settings on the left of the prompt bar).
2. "Surreal forest scene with giant mushrooms, misty atmosphere, cinematic lighting"
3. "Cyberpunk city skyline at night, neon lights, detailed buildings, ultra-realistic"

• Image Size: Use the slider to select your image (canvas) size. 1:1 is ideal for Instagram or Facebook, 9:16 for Shorts or TikToks, and 16:9 for YouTube long-form content and similar platforms.
• Aesthetics: Adjusting the values for Stylization, Weirdness, and Variety will change the types of outputs you receive—essentially, how much creative freedom you allow Midjourney to take with your prompt. You'll see in the video lectures that I prefer to leave all of these at zero, and instead focus on crafting accurate prompts to achieve the image results I want.
• Model: As a general rule, set the Version to the newest available—why not take advantage of the latest improvements?
  The difference between Raw and Standard often comes down to how strictly the model follows built-in style parameters. Raw can produce more photorealistic images by disabling those pre-set styles, but in practice, I’ve often seen little difference. Feel free to experiment and see what works best for your needs.
• More Options: Speed: Choose between Relax, Fast, and Turbo depending on how quickly you want your images generated—this is also influenced by your subscription tier. Stealth: Toggle this to control whether your image generations are visible to others or kept private.

On the left side of the prompt bar in the web version (V7), you can upload images to use in four different ways for image generation:

1. Image Prompts – Upload an image and drag it into this section to use elements of the image—such as layout, composition, or structure—in your own generation.
    
2. Style Reference – Add your image here to use it as a style reference. This influences aspects like color palette, visual style, or even a specific aesthetic (e.g., animation style, film noir, cyberpunk). Use this to replicate a particular look and feel.
    
3. Omni Reference – Use this section to upload a character reference. MJ will try to replicate that character’s likeness or appearance in your generated images.
    
4. Starting Frame – Upload an image to be used as the starting frame for a video. This allows you to generate a video that begins with your provided image and evolves from there.

Once you’ve generated an image, you can click on it to access several editing (or Creation Actions) options:

• Vary: Regenerate the image with slight or strong variations. This allows you to explore alternate versions while maintaining core elements of the original.

• Upscale: Create a higher-resolution version of the image. You can choose:
  • Subtle – Produces an upscaled version that's nearly identical to the original.
  • Creative – Upscales while also introducing slight stylistic or compositional changes.
     
• More: Re-run the same prompt or open the Edit area:
  • Edit allows you to change the canvas size (scale), and Midjourney will fill in the newly blank areas with matching background content.
  • Erase & Restore – Use the eraser tool to remove parts of the image and then prompt Midjourney to generate content specifically for that erased section.
     
• Use: This gives you quick access to:
  • Image – Reuse the image as a visual prompt.
  • Style – Reuse its style, as described in the image upload section.
  • Prompt – Re-insert the original prompt into the prompt bar for further refinement.
     
• Animate Image: Choose between:
  • Auto – Let Midjourney generate the animation based on its interpretation of your image.
  • Manual – Input your own prompt to direct the animation's motion.
     

For both options, you can choose between:

• Low Motion – Slower, more subtle camera and character movement.
• High Motion – Faster, more dynamic animations with greater scene or subject movement.

• --ar: Aspect Ratio. Adjusts the image ratio (e.g., --ar 16:9 for widescreen). NB this is in your settings on the browser version, the syntax is not needed.
• --q: Quality. Controls the render quality (e.g., --q 2 for double quality). NB you can use "upscale" in the browser version after you are satisfied with an image generation.
• --v: Version. Selects the model version for the style (e.g., --v 5). NB set this in your settings in the browser version.
• --r: Repeat (--r 4 for 4 x submissions (so 12 images total).
• --s: Style. Controls stylization level (e.g., --s 500 for a stylized look). NB this is in settings on the browser version
• --no: Negative Prompt. Excludes unwanted elements (e.g., --no people).
• --tile: Tiling. Creates seamless tile images (images you could join together seamlessly... think like a wallpaper in a house that matches and continues).
• :: 0 - 5 Prompt weight. Adds more or less weight to aspects of your prompt. This may conflict if you have "raw" or "standard" selected in your settings and instead you could adjust the "weirdness" or "stylization" (also in settings).
• --iw 0-3 When you upload an image for reference (and choose the image icon in the corner), this will adjust the image weight that Midjourney uses when referencing this image. Range 0 - 3. 
• --sw 0-1000  When you upload an image for reference (and choose the paperclip icon in the corner), this will adjust the style weight that Midjourney uses when referencing this image. Range 0 - 1000. 

For more information and to start creating, visit Midjourney's official website.

Runway ML is a browser-based AI tool for creating videos, images, and effects using simple text or image prompts. With the latest Gen-4 model, you can generate realistic or stylized videos from text , or animate still images into motion. It also supports up to 3 reference images for guiding style, characters, or layout—ideal for creators building AI-driven video content fast and easily.

Access Runway Image through the Runway platform.

Depending on when you access the tool (it changes constantly), click "Get Started" to go through to https://app.runwayml.com/video-tools/generate.

You are able to access the whole of Runway's Dashboard and suite by clicking the top-left icon on the page and then Go to Dashboard.

Gen-4 and the 2025 layout for Runway are very intuitive—being presented with the image generation tool which you can then effortlessly turn into a video.

1. Input a Detailed Text Prompt describing your desired image (more below).
    
2. References: Add up to 3 references to use them in the image generation. For example, upload characters, people, objects, and then prompt and include @reference (you can name the reference once uploaded) to include that in the image generation.
   Example: Put @reference in this scene @reference, happy, day, bright sunny, close up.
    
3. Styles: You can select a style from the provided gallery, however, I tend to prompt for this, and to get knowledge of different styles, I have added them to the /styles page you have access to.
    
4. Generate: Ensure you have selected the aspect ratio, number of images produced, and quality next to the generation button.

• Main Subject/Concept: Clearly describe the focal element.
• Details and Attributes: Specify colors, textures, lighting, and other specifics.
• Style and Mood: Indicate the desired artistic style or emotional tone.
• Technical Modifiers: Include parameters for aspect ratio, resolution, and other settings.

Example Prompts:

1. "A serene mountain landscape at sunrise, vibrant colors, photorealistic style"
2. "A futuristic city skyline at night, neon lights, cyberpunk aesthetic"
3. "A close-up portrait of a lion with a majestic mane, detailed textures, dramatic lighting"

To explore and start creating with Runway Image, visit the Runway platform.

DALL-E: AI Image Generation Tool

DALL-E, developed by OpenAI, is an advanced AI image generation tool that creates highly detailed visuals from text prompts. DALL-E is particularly strong in creating imaginative and surreal images, making it popular for unique visual concepts. The platform also includes inpainting tools, allowing users to edit parts of an image by modifying specific areas with new prompts.

1. Access DALL-E on OpenAI’s Website.
2. Input a Text Prompt that clearly describes your desired image.
3. Edit and Refine: Use the inpainting feature to modify parts of the image if needed.
4. Download and Save the final image to use in your projects.

• Subject/Theme: Describe the main subject or scene.
• Details and Textures: Specify colors, materials, and details.
• Style and Mood: Add artistic styles, lighting, or specific emotional tones.
• Perspective or Composition: Include any compositional elements like the background or angle.

Example Prompts:

1. "A vintage robot playing chess in a cozy library, warm lighting, detailed textures, realistic style"
2. "A futuristic city skyline at sunset, high detail, neon colors, cinematic perspective"
3. "Surreal underwater scene with floating castles and giant jellyfish, mystical lighting, dreamlike atmosphere"

While DALL-E doesn’t use codes like Midjourney’s (such as --ar), users can optimize prompts by carefully adjusting text details and DALL-E’s editing tools:

• Aspect Ratio: Although DALL-E doesn’t directly support custom aspect ratios like --ar, you can specify the layout in your prompt (e.g., “wide landscape format”).
• Image Quality: DALL-E automatically generates high-quality images, but phrases like “ultra-realistic” or “high detail” in your prompt can influence detail levels.
• Lighting: Specify lighting styles, such as “soft morning light” or “dramatic lighting.”
• Composition Terms: Use terms like “close-up,” “aerial view,” or “portrait format” for framing.
• Inpainting Tool: For selective edits, highlight sections to revise or add new elements.
• Art Styles: Describe styles like “oil painting,” “digital art,” or “watercolor” to influence aesthetics.
• Negative Prompts: Although there’s no direct “exclude” feature, specifying “without” or “no [element]” in your prompt helps omit unwanted items.
• Specific Artist Influence: Describe works by historical artists (e.g., “inspired by Van Gogh”).
• Time Period or Setting: Mention eras or specific settings, like “Victorian,” “sci-fi,” or “ancient Greece.”
• Mood Descriptions: Use descriptors like “mysterious,” “whimsical,” or “peaceful” to set tone.

Tips for Best Results with DALL-E: Experiment with descriptive prompts, test inpainting for refined adjustments, and add phrases that guide style and detail. DALL-E is highly responsive to clear, vivid language, helping to capture unique and complex visuals effectively.

Explore more and start creating with DALL-E on OpenAI’s DALL-E page.

Step 1: Open Gemini

Visit https://gemini.google.com/ and sign in with your Google account.
Make sure you're using a version that supports image generation (typically available through Gemini Advanced or within Google Workspace).

Step 2: Start a New Prompt

In the prompt bar, begin typing a detailed description of the image you want to generate.
Gemini uses Imagen 2, which performs best when your prompt includes clear context or narrative purpose.

Example Prompt:
Create a realistic image of a modern solar-powered home in the desert at sunset, with a Tesla parked outside and soft orange lighting reflecting off the windows.

Step 3: Include Specific Details

To improve accuracy and style, include the following in your prompt:

Purpose – What is the image for? (e.g., product mockup, educational content, storyboard)
Context – Setting, time of day, environment, or cultural background
Style – Photorealistic, flat vector, painterly, 3D render, etc.
Mood or Tone – Calm, energetic, dramatic, professional, etc.
Characters – Descriptions of people, animals, or subjects, if needed

Step 4: Submit Your Prompt

Press Enter or click Send to generate the image.
Gemini will produce one or more image options based on your input.

Step 5: Refine or Regenerate

If the result isn’t quite right, you can:

• Click "Regenerate" to try a variation
• Edit and rephrase your prompt for more specific results
   

Example Refinements:
Add fog in the background
Use a flat vector style for educational use

Step 6: Download or Use the Image

Click on the image to download it or copy it for use in your project.
Images can be used in slides, presentations, video creation, or any digital content workflow.

• Write prompts like a creative brief or story scene
• Use function or context in the prompt (e.g., "For a school science video")
• Describe the scene layout, style, and emotion
• Gemini can also create image sequences (e.g., “Show 3 steps of planting a tree”)

For more details or to start using Google Gemini, visit Google Gemini’s official page.

1. Stable Diffusion: AI Image Generation Tool

Stable Diffusion is a popular open-source AI image generation tool known for its ability to create high-quality images from detailed text prompts. Unlike some proprietary AI tools, Stable Diffusion allows users more freedom in customization and even enables local installations for those who prefer running it on their own hardware. With versatile prompt parameters, Stable Diffusion is ideal for artists, designers, and creators looking for precise control over image styles, quality, and composition.

1. Input a Detailed Text Prompt describing your image concept.
2. Add Modifiers to adjust image quality, aspect ratio, style, and other features.
3. Generate and Refine: Review your image and make adjustments as needed.
4. Save your completed image for use in your projects.

• Main Subject/Theme: Describe the central element of your image.
• Detailed Attributes: Include specific colors, textures, lighting, and additional features.
• Artistic Style: Add art styles or specific aesthetics (e.g., watercolor, photorealistic).
• Technical Modifiers: Adjust settings like quality, aspect ratio, or depth for a tailored result.

Example Prompts:

1. "A futuristic cityscape at night, neon lights reflecting on wet streets, photorealistic, cinematic wide angle"
2. "An enchanted forest with glowing mushrooms, misty background, fantasy art style, ethereal lighting"
3. "Portrait of an elderly man with deep wrinkles, soft lighting, hyper-detailed realism, 35mm film style"

• Aspect Ratio: In the description bar, select the aspect ratio icon to choose between 16:9, 3:2, 1:1, 4:5, and 9:16.
• Select a Style Image: Using the painter’s palette icon in the description bar, you can either upload an image (if you don’t have any generated images) or select a generated image in the style you wish to use for your new image. Think of this as a "style guide" for the model to follow.
• Generate Another Image Quickly: Click the regenerate icon in the top left of your image to generate a new image quickly.
• Reinvent: This option will use the structure of the original image (composition and subjects) to create a new image with updated generation.
• Reinvent with Other Settings: The three-dot icon allows you to:
• 1. Select Sketch to Image (then prompt for the style you want, e.g., "as a painting").
• 2. Generate a New Image with the Same Style (prompt what you want for the new image, and the AI model will retain the style of the original image).
• 3. Generate a New Image with the Same Structure (prompt for a new image style, e.g., "Anime," while maintaining the composition of the original image).
• Use Toolbox: The toolbox icon in your generated images allows you to directly edit the image. You can upscale, inpaint (add features), erase parts of the image, replace the background, zoom out, upscale, convert the image into a video, recolor, and more.

To start creating with Stable Diffusion, visit Stability AI’s website.

Dream Studio: AI Image Generation Tool

Dream Studio, powered by Stable Diffusion, is an intuitive, web-based tool for generating high-quality images from text prompts. Designed to be user-friendly, Dream Studio combines powerful image generation capabilities with a streamlined interface, making it accessible for users of all experience levels. With various customization options, Dream Studio is ideal for creating unique, detailed visuals quickly and easily.

1. Input a Detailed Text Prompt describing your desired image.
2. Add Customization with "Style" and "Settings"
3. Generate and Refine: Use Edit to add or remove elements
4. Download the finalized image for your project.

• Main Subject: Describe what should be in focus.
• Details and Textures: Specify colors, textures, and intricate details.
• Style and Mood: Mention artistic styles or themes to set the tone.
• Modifiers: Use codes for aspect ratio, resolution, and other parameters for precision.

Example Prompts:

1. "Majestic mountain landscape at sunrise, vibrant colors, realistic style, panoramic view"
2. "A bustling cyberpunk street with neon lights, high detail, cinematic lighting, nighttime"
3. "Close-up portrait of an owl with piercing yellow eyes, hyper-detailed feathers, dramatic lighting"

• Select your style: Enhance, Anime, Photographic, Comic book, Low poly, Clay and more
• Use Negative Prompt in settings to avoid aspects in your generation
• Upload an Image to make changes to a specific iteration you want such as style and aspect ratio
• Inside Edit select the image count you desire and add a prompt (like generative fill in Photoshop). 

To explore and start creating with Dream Studio, visit Dream Studio on Stability AI’s website.

Adobe Firefly: AI Image Generation Tool

Adobe Firefly is Adobe’s innovative AI-powered tool for generating images from text prompts, offering a creative approach that seamlessly integrates into Adobe’s ecosystem. Firefly is ideal for generating high-quality, customizable images with a range of styles and effects, and it supports powerful editing options. Designed with creators in mind, Firefly is particularly useful for professionals who work with Adobe tools, as it allows for easy integration with apps like Photoshop and Illustrator.

1. Access Firefly on Adobe’s Platform through your Adobe account.
2. Input a Detailed Text Prompt that describes your image concept.
3. Use Customization Settings to fine-tune the style, resolution, and specific elements.
4. Generate and Preview: Adjust and regenerate as needed until the image fits your vision.
5. Export or Transfer directly to other Adobe apps for further editing and enhancement.

• Main Subject/Concept: Describe the focal element or theme.
• Details and Attributes: Include colors, textures, materials, and specific attributes.
• Style and Artistic Influence: Add styles like watercolor, photorealism, or digital art.
• Technical Adjustments: Apply modifiers for aspect ratio, resolution, or lighting.

Example Prompts:

1. "A vibrant sunset over a calm ocean, watercolor style, soft brushstrokes, high detail"
2. "Futuristic cityscape at dusk with neon lights, cinematic lighting, photorealistic, 16:9 aspect ratio"
3. "Portrait of a wise old man in Renaissance attire, highly detailed, dramatic shadows, oil painting style"

• General Settings: Enable fast mode (super fast!). Here, you can also set your aspect ratio: 16:9, 1:1, 3:4, or 4:3.
• Content Type: Set to either Art or Photo.
• Composition: Use a reference by uploading an image if you have a specific composition in mind, or browse AF's gallery to select a composition.
• Styles: Set your image reference by uploading an image or browsing internally.
• Effects: Think of this as the image style. Choose from dozens of options AF offers (this means you don’t need to include these details in your written description).
• Color and Tone: Select from vibrant, muted, black and white, and more.
• Lighting: Choose from options like low light, studio lighting, dramatic lighting, and more.
• Camera Angle: Set the position from which the image is "taken."

To start creating with Adobe Firefly, visit Adobe Firefly’s official page.

Step 1: Access Meta AI Image Tools

Meta's image generation features are available through platforms like Meta AI on Messenger, Instagram, WhatsApp, and increasingly through meta.ai on desktop and mobile.

To use Meta's image generator:

• Visit https://www.meta.ai/ (if available in your region), or
   
• Use Meta AI within chat interfaces like Messenger, by typing your prompt directly into the conversation
   

Step 2: Enter a Descriptive Prompt

Meta AI is optimized for emotionally expressive, human-centered images. Start by typing a clear and detailed description of what you want to see.

Example Prompt:
Create a photo-realistic image of a woman standing in a rain-soaked city street at night, under a red umbrella, with glowing neon reflections on the pavement.

Step 3: Focus on Human and Emotional Context

Meta’s model responds especially well to prompts that include:

Characters – Descriptions of people, their age, ethnicity, clothing, or pose
Emotion – How characters should feel or express (e.g., joy, awe, mystery)
Environment – Setting, background, lighting conditions
Mood or Tone – Romantic, cinematic, dramatic, serene, etc.
Style – Photo-realistic, digital painting, impressionist, 3D render, etc.

Step 4: Generate the Image

Submit your prompt to begin the generation process.
Meta AI will return one or more images based on your description.

On platforms like Messenger, these images appear directly in the chat thread. On desktop, you’ll typically get a gallery preview.

Step 5: Refine and Iterate

You can rephrase your prompt or ask Meta AI to adjust the image:

• Add more detail to focus the composition
   
• Change lighting, style, or expression
   
• Ask for alternate versions of the same scene
   

Example Refinements:
Make the background misty and more dreamlike
Change the color palette to cool tones

Step 6: Save and Use

Once you're satisfied, download the image or save it directly from the platform.
Meta AI’s outputs can be used for concept art, video reference frames, visual storyboards, or social media content.

• Emphasize people, stories, and emotion—this is where Meta's model excels
• Use specific yet natural language (like a short film scene description)
• Include lighting, time of day, and atmosphere for cinematic results
• It supports both artistic and realistic styles, but excels at expressive realism

To learn more and start creating with Meta AI, visit Meta AI’s official page.

Grok by X (formerly Twitter): AI Image Generation Tool

Grok, developed by X (formerly Twitter), is an AI-powered tool for generating images from text descriptions. This tool uses advanced language models to transform prompts into visually compelling and diverse images, from photorealistic scenes to stylized artwork. Known for its accessibility Grok allows users to adjust parameters such as aspect ratio, quality, and style with conversational prompting, giving them control over the final output’s look and feel.

1. Access Grok through the X platform with your X account.
2. Enter a Text Prompt describing your desired image.
3. Add Modifiers to customize aspects like quality, style, and composition.
4. Generate and Review the image, and adjust prompts if needed.
5. Download the final image for your project.

• Subject or Theme: Describe the main subject or concept.
• Attributes and Details: Include colors, textures, and specific details.
• Style and Mood: Specify an art style, lighting, or atmosphere.
• Technical Modifiers: Use codes for aspect ratio, quality, and depth.

Example Prompts:

1. "A peaceful lakeside cabin at sunset, warm colors, photorealistic, --ar 16:9"
2. "Surreal forest with neon mushrooms, soft fog, dreamlike ambiance, --quality high"
3. "Portrait of a medieval knight in armor, dramatic lighting, ultra-realistic details, --style renaissance"

• Use X's own freedom of expression mandate to enable you to greate images without filter (Note, nothing illegal, derogatory, dangerous or adult).  

For more information or to start creating with Grok, visit Grok on X’s platform.