3 Insane Ways To Use Nano Banana + Kling 2.6 Motion Control For Meta Ads In 2026 (full walkthrough)

Quick Overview

The video demonstrates a workflow for generating highly customized, motion-controlled AI video ads using a combination of Nano Banana for image editing/prompting and Kling 2.6 for video generation, showing how to swap actors, change settings like clothing and environment, and repurpose existing video content for Meta Ads in minutes, not weeks.

Key Points: The core workflow involves five steps: Select Starting Video, Take Screenshot of Frame, Edit in Nano Banana, Upload video + Image to Kling, and Enjoy Your New Scene! The process leverages Nano Banana (an image editing tool) to modify a reference image (e.g., changing clothing to a Vans sweatshirt or changing the subject's age/character type) before using the modified image as input for Kling 2.6. Kling 2.6's Animate Actor feature successfully transferred motion from a source video (e.g., a dancer or person talking) onto the modified reference image, demonstrating motion control. Examples shown include swapping a person's clothing (Vans sweatshirt), changing the subject's age (to a woman in her 60s), and changing the environment (from a gym to a busy coffee shop). The presenter highlights the speed, generating multiple creative variations (e.g., 4 per prompt) in minutes, which is significantly faster than traditional methods. The speaker mentions having managed over $150,000,000 in ad spend over the last 12 years, emphasizing the value of rapid creative iteration for ad performance.

Context: The presenter, Caleb Kruse, demonstrates an advanced creative automation workflow aimed at scaling Meta Ad production quickly. This method combines two key AI tools: Nano Banana for precise image manipulation via prompting, and Kling 2.6 (a motion control AI video model) to animate the modified images based on a source video's motion. The goal is to create numerous, high-performing, targeted ad variations rapidly for direct response marketing.

Detailed Analysis

Caleb Kruse details a five-step process for rapidly generating AI-driven video ads using Nano Banana and Kling 2.6 Motion Control. First, select a starting video, then take a screenshot of the desired frame. Next, edit that image within Nano Banana using text prompts and reference images to make specific changes—such as altering clothing (Vans sweatshirt), changing the person's age (to a woman in her 60s), or changing the entire scene (to a busy coffee shop). The modified image and the original source video are then uploaded to Kling 2.6's Animate Actor feature. Kling 2.6 transfers the motion from the source video onto the newly prompted image, creating a completely new, motion-controlled video clip. Kruse shows successful applications across different scenarios, including swapping a gym influencer for a person in an orange Nike tracksuit and even swapping a real person for fictional characters like an Avatar or Mario. He emphasizes that this process allows marketers to test numerous creative variations (e.g., generating four variations per prompt) in minutes, which is crucial for scaling ad spend effectively, noting his own experience managing $150 million in ad spend over 12 years.