Parse JSON String Into Seperate Bundles

I have a claude prompt completion which is returning the following JSON string.

{
“hero_image”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A dynamic collage of social media platforms with vibrant icons, a businessperson holding a smartphone, and graphs showing upward trends. The image is set against a backdrop of interconnected network lines, captured in HDR, 8K resolution with perfect composition and intricate details.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Social media offers powering business growth, showcasing platforms, engagement, and analytics for comprehensive marketing strategy”
}
},
“image1”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A diverse group of people engaging with various social media platforms on smartphones and tablets. The devices display likes, comments, and shares. The scene is brightly lit with a modern office setting in the background, captured in HDR, 8K resolution.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Social media engagement driving business growth through diverse user interactions and content sharing”
}
},
“image2”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A split-screen image showing four different social media offers: an exclusive discount code, a giveaway announcement, an influencer collaboration, and an interactive poll. Each quadrant is vibrant and eye-catching, with clear calls-to-action, captured in HDR, 8K quality.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Compelling social media offers showcasing discounts, giveaways, influencer partnerships, and interactive content”
}
},
“image3”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A professional marketer analyzing social media advertising dashboards on multiple screens. The screens display targeting options, audience demographics, and performance metrics. The scene is set in a modern office with social media icons floating in the background, captured in HDR, 8K resolution.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Social media advertising strategy leveraging targeted reach and analytics for business growth”
}
},
“image4”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A montage of various video content formats: a live stream, a product tutorial, behind-the-scenes footage, and a short-form video. Each video is displayed on different devices, surrounded by engagement icons. The scene is dynamic and colorful, captured in HDR, 8K quality with perfect composition.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Video content dominating social media marketing with diverse formats enhancing engagement and reach”
}
},
“image5”: {
“version”: “39b3434f194f87a900d1bc2b6d4b983e90f0dde1d5022c27b52c143d670758fa”,
“input”: {
“prompt”: “A futuristic social media command center with holographic displays showing emerging platforms, AI-driven analytics, and trend forecasts. A diverse team of marketers interacts with the displays, surrounded by floating social media icons and data visualizations. The scene is captured in HDR, 8K resolution with intricate details.”,
“aspect_ratio”: “1:1”,
“output_quality”: 100,
“alt_text”: “Future of social media marketing featuring AI integration, emerging platforms, and innovative strategies”
}
}
}

When I then use the Parse JSON module I get one collection bundle when I would like the JSON string to split into 6 separate bundles for each image prompt collection.

I’ll upload blueprint with all the data.

I’ve tried searching the community but can’t find the solution to solve my issue.

Any help would be great.

Thank you

blueprint.json (7.7 KB)

Your data structure isn’t set right, i’d ask you to get claude to output the data as an array of images.

It needs to look similar to this to get it to work

@D-EFFCON - Thank you! made the changes and it worked a treat… sometimes you need to take a step back. Have a great day and thanks again :star_struck: