Generating Images and Videos

The assistant can generate images and short videos for you on demand. Both happen inside the conversation, no separate tool required.

Generating images

Ask for an image in plain language:

Generate a hero image for a blog post about sustainable cities. Minimalist, illustration style, soft greens and blues.

The assistant generates between one and four images, depending on what you ask for. Each one appears in the chat as a thumbnail.

You can pick a style:

• Photorealistic — Looks like a photograph.
• Illustration — Hand-drawn or vector look.
• Minimalist — Simple shapes, lots of negative space.
• Abstract — Non-representational.
• Infographic — Diagrammatic look with iconography.

You can:

• Click an image to view it full size.
• Download it to your device.
• Copy it to your clipboard.
• Ask the assistant to regenerate with a tweak — "Make it darker", "Replace the buildings with trees".

A note on text in images

The image generator does not embed text. If you need text on top of an image — a heading, a caption — generate the image and then add the text in a presentation, document, or external tool.

Generating videos

Ask for a short video in the same conversational style:

Generate a 5-second video of a coffee cup being placed on a wooden table, soft morning light, slow zoom-in.

Video generation takes longer than image generation — usually a few minutes. The assistant will show you a progress step while it works. When it's ready you'll get a playable video element in the chat.

Videos:

• Are stored securely in your account.
• Can be downloaded from the chat.
• Are short clips, not long-form content.

Tips for better results

• Be visual. Describe lighting, mood, perspective, palette.
• Be specific about the subject. "A golden retriever puppy" is better than "a dog".
• Ask for variations. "Show me three different angles" gives the assistant room to explore.
• Iterate. If something isn't quite right, describe the change and ask it to regenerate.

See also

• Tools the Assistant Uses — the broader tool set.
• Presentations — where you'll typically combine generated images with text.