Kling AI: Using multiple visual references with Kling 3.0 Omni
Kling AIDemo summary
The creator shows how to upload up to seven different image references (characters and locations) and tag them in the prompt to guide the AI's generation.
Step-by-step
- Select Kling 3.0 Omni in the reference settings
- Upload up to seven reference images from your local files or previous generations
- Upload multiple angles (front, side, back) of a single character to provide more data
- Tag the images in your prompt using their assigned numbers (e.g., 'Image 1 and Image 2')
- Describe the interaction or story between the tagged images and the location reference
Options
- Choose references from your own previous generations
- Upload any type of image from your computer
- Use multi-shot mode or standard prompting
Watch out for
- The system can still glitch or produce inconsistencies occasionally
- You must manually tag the image numbers in the prompt for the AI to associate them correctly
Tips
- Provide three different angles of a character to prevent the AI from making up random details
- Use a specific prompt to generate a character sheet with multiple angles rather than manually editing them in Photoshop
- Refer to characters and locations by their image number to maintain consistency in the story
Highlights
“this is super powerful if you have all kind of little details that need to be there in your shot”
All demos from “STOP Wasting Credits & Master Kling 3.0 in 25 Minutes”
- 1:201:09Configuring Kling 3.0 Multi-shot settingsThe creator demonstrates how to select the Kling 3.0 model and enable the multi-shot feature to customize up to six different shots within a single video generation.Kling AI· AI Movie Generator
- 3:321:08Generating consistent character angles with Multi-shotUsing a single image reference, the creator shows how to describe multiple camera angles and actions for a character to generate a consistent 12-second multi-scene video.Kling AI· Image to Video
- 7:060:58Text-to-video generation in OpenArtThe creator demonstrates the text-to-video workflow by inputting a descriptive prompt of rugby players in a locker room to generate a scene from scratch.OpenArt· Text to Video
- 9:382:18Using multiple visual references with Kling 3.0 OmniCurrentThe creator shows how to upload up to seven different image references (characters and locations) and tag them in the prompt to guide the AI's generation.Kling AI· AI Image Generator
- 15:001:06Image-to-video with start and end framesThe creator demonstrates the image-to-video method by setting a specific start frame and using multi-shot prompts to control the motion and narrative flow.Kling AI· Image to Video
- 17:150:29Extending video duration with Start/End frame interpolationThe creator shows how to use a start frame and a different end frame while sliding the duration bar to generate a longer, continuous video sequence.Kling AI· AI Video Extender
- Watch “STOP Wasting Credits & Master Kling 3.0 in 25 Minutes” →
AI Image Generator
- 9:382:18Using multiple visual references with Kling 3.0 OmniCurrentThe creator shows how to upload up to seven different image references (characters and locations) and tag them in the prompt to guide the AI's generation.Dan Kieft