Image-to-Video Generation

Image-to-video generation lets an Agent One agent turn one or more still images into a short MP4 video from a written prompt. Use it for product teasers, social clips, concept previews, animated mockups, and visual variations based on images you provide.

Availability

Image-to-video generation is available on the Basic plan and higher.

Some public showcase agents and free marketing tools do not generate videos, even when they can answer questions about the product. If a user tries to generate a video in a place where the feature is not available, the chat will show an upgrade or availability message instead of starting a render.

What You Can Provide

You can give the agent:

• One or more uploaded images
• A prompt that describes the video
• Optional direction for style, duration, music, voiceover, or resolution

For best results, upload clear images that are relevant to the same scene, product, or idea. The agent can use multiple images together when your prompt asks for a sequence, comparison, transformation, or montage.

Creating a Video in Chat

1. Open a chat with your agent.
2. Click the attachment button and upload the images you want to use.
3. Wait for the upload tiles to finish. Uploaded items show as ready; failed uploads can be retried or removed.
4. Send a prompt that asks for a video.
5. The chat will show a video status card while the video is being generated.
6. When the video is ready, it appears inline with video controls and a download option.

Example prompt:

Turn these product images into a 10 second social ad. Start with a close-up, add slow camera movement, keep the mood premium and minimal, and end on the product against a clean background.

Writing Better Prompts

A strong video prompt usually includes:

• Subject: What should be shown in the video
• Motion: How the image should move or change
• Mood: The style, energy, or emotional tone
• Format: Length, pacing, or destination such as social ad or product teaser
• Constraints: Anything to avoid, such as changing the product shape or adding extra text

Good prompts are specific without being overloaded. If the first result is close but not exact, ask the agent to revise with one or two clear changes.

Upload Feedback

The chat input shows a small preview strip for uploads:

• Image uploads show thumbnails.
• Document uploads show file-style previews where supported.
• Each upload shows progress while it is running.
• Failed uploads show a failure state with a retry control.
• You can remove an upload before sending the message.

On small screens, the preview strip stays compact so the chat input remains usable.

Troubleshooting

I see an upgrade prompt

The agent owner must be on the Basic plan or higher to generate videos. Upgrade the account that owns the agent, then try again.

The feature is unavailable in a public demo

Some public demo agents are designed as free product previews and do not include video rendering. Create or open your own agent on an eligible plan to use image-to-video generation.

An upload failed

Retry the upload from the preview tile. If it fails again, check the file type, file size, and network connection, then upload a fresh copy.

The video card appears but the video is not ready yet

Video generation can take longer than a normal chat response. Leave the chat open while the video status card updates. If the video does not appear after several minutes, try refreshing the thread or asking the agent to generate a new version.

The result does not match the prompt

Try again with tighter direction. Mention the exact subject, motion, camera style, text restrictions, and what should stay unchanged from the source images.