Videos

Generate professional video clips from text, or create presenter videos with talking avatars — then publish to your platforms or share with a public link.

What you can create

Videos turns ideas into finished video content. Describe a scene and the AI generates a video clip, or write a script and a presenter delivers it on camera. Every generated video is saved to a personal library, ready to publish or share.

Two generation modes cover different use cases:

Scene generation — Turn a text description or reference image into a video clip. Ideal for promotional content, product demos, social posts, and visual storytelling.
AI Presenter — Write a script, choose an avatar and voice, and get a video of a presenter speaking the content. Build multi-scene presentations for training materials, explainers, and announcements.

Supported AI providers

Video generation works with any connected video provider:

Scene generation

Google Veo, Kling, Runway, MiniMax, Luma Labs.

AI Presenter

HeyGen, D-ID, Synthesia.

Your own API keys

Video generation uses connected provider keys — each generation is billed directly by the provider. Cost details appear in usage history. Connect at least one video provider in AI Settings to get started.

Generating a scene video

Select Scene mode on the Generate tab.
Describe the video. Include visual details, camera movement, lighting, and mood. Example: "Cinematic aerial shot of a Tuscan villa at golden hour, camera slowly orbiting, warm lighting."
Pick an orientation — Landscape, Square, or Portrait.
Select an AI model from the dropdown.
Fine-tune settings (optional) — Expand Video settings for duration, resolution, camera movement, and audio.
Add reference images (optional) — When the selected model supports it, attach images from the Images library to guide the visual style.
Click Generate video. A progress notification stays visible as you navigate to other pages. Up to 3 videos can generate simultaneously.

When a video is ready, the notification updates with View and Download actions.

Creating a presenter video

Select Presenter mode on the Generate tab.
Choose an AI provider — HeyGen, D-ID, or Synthesia.
Choose a presenter. Pick a stock avatar from the vendor, upload a photo, or use an image from the Images library.
Configure voice — Select language, voice, speed, pitch, and emotion. Use Preview voice to hear a sample.
Write the script. Keep it conversational for natural delivery.
Set background and orientation. Pick a background color and an aspect ratio.
Click Generate video. Progress tracking works the same as scene videos.

Multi-scene presentations

For providers that support it, build videos with multiple scenes — up to 20 per video — each with its own script and settings. Click Add scene to add scenes, write separate scripts, drag to reorder, or use Duplicate to start from an existing one. The final video combines all scenes in order into a single output.

Managing videos

The My videos tab shows every generated video. Per-card actions: view (play inline), download, publish to connected platforms, share publicly, or delete.

Publishing to platforms

Publish any video directly from the library to connected platforms. WebGPT generates AI-powered metadata — title, description, and tags — optimized for each destination. Review and edit metadata before publishing. Publish to multiple platforms in one step.

Create a public share link for any video. The shared page has an embedded player and basic video details — anyone with the link can watch without logging in. Revoke the link any time to remove public access.

Usage

Each generation counts as one use against the daily video limit on your plan. The usage indicator on the page shows the remaining daily allowance. See Plans & usage limits.

We use cookies

Menu