The Quick Workflows
There are several ways to achieve this, but in the end there are a couple basic flows to consider:
Character Reference: Create and train a model with a specific character to use directly in prompts.
Image to Video: Use images to guide the video creations, you can create images with the consistent character or product.
One of the strongest tools right now to achieve this is Kling 3.0, Veo 3.1 and Google Nano Banana Pro.
The below post is a continuous video that is extended from one 15 second clip.
It captures the last shot of the first video and uses it as the first frame of the next video.
This gives it that continuation feeling, and Kling 3.0 manages to maintain the consistency by following the text prompt closely. The character remains the same.
This is started as a single image from Midjourney, which guides the output.
Let’s dive in a bit deeper…
The Deeper Dive
KLING 3.0 ELEMENTS
With Kling 3.0 Omni, you can design a character from a single image. This feature, called “elements”, also works with products too.
From providing a voice (through uploading an MP3 to use in dialogue), you are also able to include the element in a prompt. This aids the AI video tool as it can directly draw from a mini-trained model based on your image.
Kling 3.0 Omni Elements
What’s best about the element feature is it encourages shots from different angles as well to help keep the consistency in the result. The videos can now show your character turning and moving whilst keeping the face more consistent.
GOOGLE NANO BANANA PRO
Google Nano Banana Pro is the most accessible and easiest tool to use for image consistency and modification.
Accessible directly on Gemini itself, or on other tools such as Freepik, you can upload an image of a character or product as a reference to be included into a shot.
Nano Banana Pro with Freepik
When you create the images, or storyboard, for a character, you can then bring it to the AI video step using image-to-video flows.
Combined with a text prompt and you can get those longer videos that keep the character but across different scenes.
You retain the creative control by always knowing the first frame, and then you have to experiment and test with the text prompt after.
Here is a powerful example of an extended video with Kling 3.0 that involves a character reference!
START AND END FRAMES
Video tools like Google Veo 3.1 give you the ability to upload a start and end frame.
With this level of control all you have to worry about are the two images, and how to control what happens in between.
Google Nano Banana Pro and Veo 3.1 on Freepik
Veo 3.1 also gives you character dialogue options which you can add to your prompt for native audio.
Create your first image of the character, edit it with Nano Banana Pro to make a change, and then place these as the start and end scene.
You’ll retain the same character, same environment but with a completed video.
AI TOOL DISCOUNTS
Don’t pay full price, use these codes:
Freepik: https://referral.freepik.com/mQI968S
Runway Code: JERROD25
