Video Scenes and Clips

ImagiPortal offers two different video experiences:

Use a video scene when you want to see and hear an AI companion deliver a message. Use a video clip when movement, action, atmosphere, and the sound of the scene are more important than synchronized speech.

Create videos from a conversation

Conversation tools use a companion's response and the recent conversation as their starting context. The result stays connected to the selected message.

Open Tools under an AI companion's message to find Generate video scene and Generate video clip.

Generate a talking video scene

A video scene brings a spoken response to life. The character speaks with lip movement, facial expressions, gestures, and natural body movement that follow the generated audio.

The selected message needs two things:

• An image: the visual starting point for the scene
• Spoken audio: the companion's message spoken with the character's active voice

If either one is missing, ImagiPortal shows what is required and lets you generate it first.

Create the scene

1. Open a conversation with an AI companion.
2. Find the companion message you want to turn into a scene.
3. Select Tools under the message.
4. Select Generate video scene.
5. If asked, generate the missing image or spoken audio.
6. Select Generate video scene again when both are ready.
7. Review the image, processing information, and access requirement.
8. Confirm the generation.

The finished scene appears under the same message.

How the scene is created

ImagiPortal uses the generated image as the character's appearance and the spoken message as the performance. The character's mouth and movement are synchronized with the audio.

The scene length follows the length of the spoken message. A longer voice message creates a longer scene and can require more processing time.

Choose a clear, front-facing image when possible. A visible face and upper body give the video more useful information for speech, expression, and gestures.

Read Voice Studio to learn how the character's active voice is selected.

Generate a video clip from a conversation

A video clip adds motion to an image connected to a companion's message. It can animate the character, camera, background, clothing, weather, or another visual part of the scene.

A clip needs an image but does not use the spoken version of the companion's message. Instead, it includes sound generated for the scene, such as ambience and sounds connected to the visible action.

For example, a rainy street clip may include rain and city sounds. A beach clip may include waves and wind. This scene sound is different from a video scene, where the character speaks the selected message and the mouth movement follows that speech.

Create the clip

1. Open a conversation with an AI companion.
2. Find a companion message that has a generated image.
3. Select Tools under the message.
4. Select Generate video clip.
5. If the message has no image, choose to generate one first.
6. Return to Tools and select Generate video clip again.
7. Review the starting image and generation information.
8. Confirm the generation.

The conversation clip is 10 seconds long and appears under the selected message when it is ready.

ImagiPortal prepares the movement prompt from the selected message, its image description, and recent conversation context. This helps the motion stay connected to the current scene.

Video scene or video clip?

Choose based on what matters in the result:

Create a clip in the Video Generator

Use the standalone Video Generator when you want to write your own video prompt or begin with an image outside a conversation.

1. Open Generator.
2. Select Video.
3. Choose whether the video uses an AI character.
4. Add a reference image when needed.
5. Describe what should happen.
6. Select Generate.

The Video Generator creates short video clips with generated scene sound. It does not create the synchronized talking scenes described above. Talking video scenes begin from a companion message in a conversation.

Text to video

Text to video creates a new clip from your written description without using a starting image.

1. In the Video Generator, select No Character.
2. Leave the reference image empty.
3. Describe the subject, action, setting, camera, and mood.
4. Select Generate.

Because there is no reference image, the video creates the full visual scene from the prompt. Use this option when you want a general video that does not need to preserve an existing character or image.

The result appears in Generic Videos.

Image to video without a character

Use an uploaded image when you want to animate your own visual starting point.

1. Select No Character.
2. Open the image selector.
3. Upload a reference image.
4. Describe the movement you want to see.
5. Select Generate.

Uploaded images must be smaller than 10 MB. ImagiPortal checks the image before generation. Use only an image that you have the right to upload.

The image shape determines whether the video uses a wide or vertical format. The image is centered and may be cropped to fit the video frame.

Create a video with an AI character

Use the character workflow when an existing AI character should appear in the clip.

1. Open the character selector.
2. Choose My Characters or Public Characters.
3. Search for and select a character.
4. Select one of the character's available images.
5. Describe the movement, action, camera, and atmosphere.
6. Select Generate.

A reference image is required when a character is selected. The image helps preserve the character's appearance while the prompt explains what should move or happen.

The result appears in that character's video collection.

Write a useful video prompt

Focus on movement and keep the direction clear. Include:

• Subject movement: walking, turning, smiling, dancing, or looking around
• Camera movement: slow zoom, gentle pan, tracking shot, or still camera
• Environment movement: rain, wind, waves, traffic, smoke, or moving lights
• Scene sound: ambience and action sounds that should fit the video
• Mood and pace: calm, energetic, dramatic, romantic, or mysterious
• Important limits: details that should remain still or unchanged

For example:

She turns toward the window and smiles. Rain moves across the glass while the camera slowly moves closer. Her movements are calm and natural.

Avoid adding too many unrelated actions to one short clip. One clear movement often produces a more stable result than a long sequence of events.

Choose a useful reference image

Video generation begins from the details already present in the image. For a more stable result:

• Use a clear, high-quality image
• Keep the main subject easy to see
• Avoid heavily blurred faces or hands
• Choose an image with enough space for the intended movement
• Match the image shape to the video you want when possible
• Start with a strong character image before asking for complex motion

The video interprets the source image rather than preserving every detail exactly. Faces, clothing, hands, backgrounds, and small objects can change while the image is animated.

While a video is generating

Video generation can take longer than image generation. The app shows progress or an estimate when that information is available.

You can continue using ImagiPortal while the video is processing. When a result takes longer than expected, the app can notify you after it is ready.

Find and manage generated videos

Videos created from conversations remain attached to their original messages. You can also find character-based conversation videos in Your Media Gallery on the character's main profile.

The Video Generator organizes standalone results into:

• Generic Videos for results without a character
• Character video collections for results connected to a character

Depending on the gallery and device, available actions can include:

• Play the video at a larger size
• Download it
• Copy its prompt
• Generate another version
• Delete it

Deleting a conversation video removes that generated video from the message. It does not delete the text message, its image, or its spoken audio.

Access and pricing

Video generation is a paid feature and can use PORTI even with an active subscription. The app shows the current requirement before generation.

See Plans and Subscriptions for current pricing and access information.