Generating Midjourney images using Make.com

The speaker, Gizm from imagineapi.dev, introduces the goal of integrating Imagine API with make.com to create image generations without coding. They plan to use Typeform to create a form called 'Freaky Friday' that will collect user prompts and emails. The process is experimental, and the speaker intends to share all the steps, including potential mistakes.

The speaker creates a Typeform trigger that activates when a new response is submitted. They connect it to a specific form and set it to only watch completed responses. The form collects a 'prompt' and an email address from the user. The speaker then adds another app to send the collected data to the Imagine API.

The speaker describes the process of sending a POST request to the Imagine API. They discuss adding a header for authentication using a Bearer token and constructing the JSON body with the prompt from the Typeform response. The speaker experiences some nervousness and uncertainty but remains committed to solving the task.

After sending the request, the speaker checks the API response and encounters a 404 error, indicating that the request failed. They go through the process of diagnosing the issue, checking the request content, and eventually identify that the problem lies in the URL structure. The speaker corrects the mistake and reruns the module.

The speaker addresses an authentication error (403) and corrects the headers by adding the correct authorization token. They then discuss setting up a repeater module to check the status of the image generation periodically. The goal is to keep asking the API about the image's progress until it's generated.

Once the image is confirmed to be generated, the speaker outlines the steps to send an email with the image to the user. They detail how to use the image ID to make a GET request to the Imagine API to retrieve the image and then use the email address from the form to send the image via email.

The speaker encounters difficulties with setting up a filter to route the response correctly based on the status. They express frustration but eventually set up a filter that checks if the status is 'completed' before proceeding to send the email. If not, the process is set to repeat.

The speaker discusses using a repeater module to automate the process of checking the image generation status. They mention the need for error handling and the importance of repeating the request until the image is successfully generated. They also touch on the concept of rollback in case of errors.

The speaker decides to adjust the timing of the repeater module to wait for 60 seconds before checking if the image has been generated. They acknowledge that ideally, the system should check every few seconds rather than waiting for a fixed duration.

After the image is generated, the speaker demonstrates sending an email with the image attached. They mention encountering an error related to the email account but manage to resolve it. The speaker also discusses the need for further refinements in the process, such as implementing a timer for continuous checks and handling potential errors.

The speaker concludes the video by showing the successful outcome of their efforts, an email with the generated image. They highlight the need for further improvements, such as implementing a timer for continuous checks and error handling. The speaker invites viewers to get a demo account to try out the Imagine API by emailing the provided address.