How To Use Google VEO 3 JSON Prompting To Create Ad Videos Like a PRO

Summary

This video tutorial by Charles demonstrates how to create viral AI-generated ads using a method called JSON prompting. It walks viewers through setting up tools like Flow V3, creating projects, and generating optimized prompts that guide AI scene-by-scene, including visuals, lighting, camera movements, and branding. Charles shows examples for brands like Nike, Lamborghini, YouTube studios, and food products, as well as animated 2D and 3D styles. The tutorial emphasizes the two-step process of optimizing descriptions and converting them into JSON prompts, allowing anyone to produce professional, cinematic, or animated ads quickly, even monetizing them for businesses.

Takeaways

• 😀 JSON prompting is a structured way to instruct AI for creating high-quality videos, not a magical shortcut.
• 😀 Properly structured prompts are essential; vague prompts often yield unsatisfactory results.
• 😀 Users should set up Germany V3, configure settings to V3, and use text-to-video with output set to one.
• 😀 Creating a JSON prompt project in Flow allows for reusable templates and detailed instructions for ad generation.
• 😀 The ad creation process involves a two-step approach: first generate a descriptive prompt, then convert it into a JSON prompt.
• 😀 JSON prompts include detailed scene-by-scene instructions covering visual props, camera movements, lighting, styling, branding, and product placement.
• 😀 Multiple video styles are supported, including cinematic live-action, 2D and 3D animation, and product unboxing.
• 😀 Example ad creations include Nike, Lamborghini, kitchen setups, YouTube studio setups, burgers, popcorn, chocolate, and Coca-Cola.
• 😀 The prompts, templates, and instructions are shared via a Telegram channel for easy access and replication.
• 😀 This method enables fast, professional-quality ad creation that can be monetized by offering services to brands or for personal business use.
• 😀 Video quality can be enhanced using upscaling to 1080p, providing polished outputs suitable for commercial use.
• 😀 Leveraging JSON prompting can save significant time and resources compared to traditional ad video production.

Q & A

What is JSON prompting and how is it used in AI ad creation?

JSON prompting is a structured method of instructing AI to create content by giving detailed scene-by-scene instructions. It is used to control elements like visuals, camera movements, lighting, styling, branding, and product placement in AI-generated ads.

Why is JSON prompting not considered a magical tool for making videos?

JSON prompting is not magical because the quality of the video still depends on how well the prompts are structured. Poorly constructed prompts can result in low-quality videos, so careful instruction is still required.

What initial steps are required to use Germany V3 and Flow for AI ad creation?

You need to search for Germany V3, access Flow, sign up for the free trial, ensure the tool is set to version V3, select output 1 to save credits, and choose text-to-video mode.

What is the purpose of creating a project in Charg?

Creating a project in Charg allows you to centralize all instructions and JSON templates. This setup enables you to quickly generate ads by reusing templates and instructions without rewriting long prompts each time.

Explain the two-step process for generating JSON prompts for ads.

Step 1: Generate an optimized descriptive prompt by providing a short idea of the ad to AI, which expands and improves it. Step 2: Convert the optimized description into a JSON prompt within the Charg project, including details like style, camera, lighting, motion, and text.

How can you generate different styles of videos, such as 3D or 2D animated videos?

You need to mention the desired style explicitly in the descriptive prompt, e.g., 'animated cartoon style' or '2D cartoon style.' The AI will then create a JSON prompt that reflects this style when generating the video.

What are some examples of ads that can be created using this method?

Examples include cinematic Nike ads where shoes form from raw elements, Lamborghini ads where cars assemble from components, kitchen unboxing videos, YouTube studio setups, burger ads, popcorn ads, chocolate ads, and Coca-Cola ads.

What is the benefit of using templates and instructions from the Telegram channel?

Templates and instructions save time and reduce effort, allowing users to quickly generate professional-quality JSON prompts without writing detailed instructions from scratch.

How can this workflow be monetized for business purposes?

Users can offer AI video creation services to brands and businesses, potentially earning $500–$1,000 per video, since the niche is largely untapped and professional-quality ads can be generated quickly.

Why is it recommended to open a new chat outside the JSON project when generating descriptive prompts?

Opening a new chat ensures the AI generates an optimized descriptive prompt rather than automatically producing a JSON prompt. This two-step process allows better control over prompt quality and flexibility.

What settings in Flow are important to conserve credits while generating videos?

Setting the output to 1 ensures that only one video is generated per request, which helps conserve credits. Additionally, users can optionally upscale the video to 1080p for better quality after initial generation.

Can this JSON prompting method be applied to both realistic and animated videos?

Yes, the method works for realistic videos as well as 3D and 2D animated videos, as long as the prompt clearly specifies the desired animation style.