Image Generation

Image generation lets you turn a conversation or a written idea into a new picture. You can generate an image directly from a companion's message or use the Generator for more control.

Generate an image from a conversation

This is the simplest way to create an image that matches the current conversation or roleplay.

Choose the image model first

Before generating, open the chat settings and find Image Model. Choose:

• True for the standard image model
• Wild for the experimental image model

The selected model is used for new images generated from the conversation. You can switch it again before generating another image.

Generate the image

1. Open a conversation with an AI companion.
2. Find the companion message you want to turn into an image.
3. Select Tools under the message.
4. Select Generate image.
5. Wait while the image is prepared and generated.

The image appears below the selected message when it is ready.

How the scene is created

ImagiPortal uses the selected message, recent conversation context, and the AI character's visual identity to prepare the scene.

This means you do not need to write a separate image prompt. A clear message that describes the place, action, clothing, or mood gives the image generator more useful context.

For example, instead of a message that only says:

She looks outside.

A more descriptive scene could say:

She steps onto the rainy balcony in her red evening dress and looks across the lights of the city.

The second version gives the image generator a clearer setting, outfit, action, and atmosphere.

While the image is generating

The image remains connected to its conversation message. The app shows its current progress while it is waiting or processing.

You can continue using the app while the image is being created. If generation takes longer than expected, ImagiPortal can notify you when the result is ready.

Work with a conversation image

Select a completed image to open it at a larger size. Depending on your device, available actions include:

• Download or share the image
• Generate a new version
• Use the image as the starting point for a video
• Delete the image

Generating a new version uses the conversation context again, so the composition and details may change.

Deleting the image removes it from the conversation and its connected Media Gallery. It does not delete the text message.

Find images created with a character

You can return to generated images without searching through old messages.

From the character profile

Open the character's main profile page and find Your Media Gallery. This gallery brings together the images and videos you have generated with that specific character.

Select the Images view to browse the character's generated pictures. Open a result to view or manage it.

From the current chat

Open Gallery from the chat and select Images. This gallery shows images generated in that particular conversation.

Use the character profile gallery when you want media from the character across your interactions. Use the chat gallery when you only want media from the conversation that is currently open.

Create images in the Generator

Open the Generator when you want to begin with your own prompt instead of a companion's message.

The Image area includes four workflows:

• Image: Create a general image without an AI character
• Character Image: Create a new scene with a selected AI character
• Selfie: Create an image of yourself with an AI companion
• Try-On: Try clothing on an uploaded photo

This chapter covers Image and Character Image. Read Selfies with an AI Companion and Try-On and Dressing Room for the personal-photo workflows.

Create a general image

Use the Image workflow when the picture does not need an ImagiPortal character.

1. Open the Generator and select Image.
2. Choose the Image workflow.
3. Select No Character.
4. Choose a style: Realistic, Anime, or None.
5. Choose an image model.
6. Describe the image you want.
7. Select Generate.

The result is saved in Generic Images.

Create a character image

Use Character Image when an existing AI character should appear in the result.

1. Open the Generator and select Image.
2. Choose Character Image.
3. Select Public Characters or My Characters.
4. Search for and select the character.
5. Choose an image model.
6. Describe the new scene.
7. Select Generate.

The selected character's identity and visual information help guide the result. Your prompt should focus on what is new in this image, such as the action, clothing, location, framing, lighting, or mood.

The image is saved in that character's collection.

Choose an image model

The Image Generator offers two model choices:

Use True as the regular starting point. Try Wild when you want to explore a different result and are comfortable using an experimental option.

Your image-model setting is also used when you generate images from conversations. You can change it from the chat settings or from the Generator controls before starting a new image.

Write a useful image prompt

You can write in simple, natural language. Include the details that matter most:

• Subject: who or what should appear
• Action: what is happening
• Location: where the scene takes place
• Clothing: what the subject is wearing
• Framing: portrait, full body, close-up, or wide scene
• Lighting: daylight, candlelight, neon light, or another clear choice
• Mood: calm, joyful, mysterious, dramatic, or playful

For example:

A full-body image of the character walking through a quiet futuristic market at night, wearing a silver jacket, with soft blue neon light and a calm mood.

Keep the main idea clear. Too many conflicting locations, outfits, actions, or visual styles can make the result less predictable.

Manage Generator images

The Generator organizes images into character collections and separate collections for generic images, selfies, and saved try-ons.

From an image collection, you can:

• Open an image at a larger size
• Copy its prompt
• Download or share it
• Use it to generate a video
• Delete it when deletion is available
• Load older results

Read Video Scenes and Clips to learn how to animate a generated image.

Character consistency

Character images use the selected AI character as their visual foundation. A clear character profile, stable avatar, appearance description, visual style, and world help different images feel connected.

If the result does not look close enough to the character:

1. Make the prompt shorter and clearer.
2. Remove conflicting appearance details.
3. Check that the character's avatar and written appearance match.
4. Generate another version.
5. Try the other image model if you want a different interpretation.

Read Avatars and World for the character settings that guide visual media.

Safety and access

Prompts and conversation context are checked before image generation. Content that does not follow ImagiPortal's rules may be blocked.

Image generation is a paid feature for regular users and is included with an active subscription. The app shows the current requirement before generation.

See Content Guidelines and Plans and Subscriptions for more information.