Automated Midjourney images using API!

In this section, Neil introduces the concept of using a third-party API to automate image generation with MidJourney, without manually interacting with Discord. He explains how this approach is useful for automating tasks like creating blog posts, social media ads, or other content that requires images. Neil contrasts the quality of OpenAI’s API with MidJourney, pointing out that the latter produces better results. The key tools for this automation are Make.com, Airtable, and User API. Neil will demonstrate how to use them step-by-step for programmatic image generation.

Neil dives into the setup for the first automation scenario, explaining how to create an image prompt using the User API. He outlines how to pass a manually created prompt (which could be dynamically generated in a real workflow) and use it to generate images via the API. He discusses how to set up a webhook URL that receives the generated image from User API and the importance of connecting the API key from User API for this process to work seamlessly.

In this section, Neil explains the setup for the second scenario, which handles the webhook response from the first step. He shows how to create a webhook in Make.com that will receive the image data generated by User API. This involves copying and pasting the webhook URL into the correct place and ensuring that it can handle the incoming data properly. Neil also prepares for the next stage by setting up placeholders to dynamically capture relevant information from the response, such as the image hash.

Neil describes the third scenario where the image from the second webhook is upscaled using the User API’s upscaling module. He explains how to set up this automation to upscale each of the four images generated by MidJourney. He adds a repeater module to run the upscaling process four times, once for each image, and dynamically inserts the correct image position in each iteration. After some troubleshooting with his User API account, he hardcodes certain values to finalize the setup and ensure everything works smoothly.

Neil explains how to run the full automation workflow from start to finish. He demonstrates how the automation generates the image, sends the response to the second webhook, and then upscales the images before uploading them to Airtable. The final images are stored in an Airtable table, ready to be used. He also highlights how this process can be dynamically customized, for example by feeding a prompt from an article or blog post. Neil concludes by suggesting the possibility of expanding the automation in future videos, such as building workflows for Facebook ads or complete blog posts.