Generating videos

TalkCut offers three approaches to video generation, each suited to different creative needs. All generated videos appear on your canvas.

​

Text to video

Describe a scene in words and the AI generates a video clip. Examples:

• “A drone shot flying over a mountain range at sunrise”
• “A cat walking across a piano keyboard, close-up”
• “Abstract liquid metal flowing and morphing shapes, slow motion”

Text to video supports all 10 aspect ratios (21:9 through 9:16) and works best for establishing shots, abstract sequences, and scenes where you want the AI to interpret your description freely.

​

First and last frame

Define the start and end of a shot by providing two images. The AI generates the transition between them — creating smooth camera movement, character action, or environmental change. To use this:

1. Place two images on the canvas — your start frame and your end frame
2. Tell the AI: “Create a video transitioning from [first image] to [last image]”

This gives you precise control over the beginning and end of a shot. The AI handles the in-between frames. Use cases:

• Scene transitions (day to night, empty room to furnished)
• Controlled camera movement between two compositions
• Character action (standing to sitting, calm to surprised)
• Environmental changes (dry landscape to rainy, calm sea to storm)

Aspect ratios: 16:9 (widescreen) or 9:16 (vertical). Duration: 5-8 seconds.

​

Reference video

Provide 1-3 reference images and a text prompt — the AI generates video that maintains visual consistency with your references. This is ideal when you need characters, environments, or props to look the same across multiple shots. Examples:

• “Generate a shot of this character walking through a city at night” (with character reference image selected)
• “Create a scene in this environment with camera slowly pushing in” (with environment reference)
• “Animate this character turning to face the camera” (with character reference)

Aspect ratio: 16:9 (widescreen). Resolution: 720p (default) or 1080p. Optional AI-generated audio. See Reference video for a detailed guide on building continuity across shots.

​

Output format

Generated videos are delivered as MP4 files.

​

Downloading videos

Click any video on the canvas and use the Download button in the toolbar to save it as an MP4 file.

​

Tips for better results

• Be specific about camera work — specify dolly, pan, tilt, rack focus, or establishing shot
• Describe lighting and mood — “warm golden hour” or “harsh overhead fluorescent”
• Use reference video when you need visual consistency across shots
• First and last frame gives you the most control over camera movement
• Start with images — generate a still image you like, then use it as a reference or keyframe for video

​

Credit costs

Video generation uses more credits than image generation because of the computational complexity involved.