Slideshow Video from Sheets Bot (ML)

**In addition to these instructions, please also refer to the NOTES in the blueprint after upload 

Slideshow Video From Sheets Midjourney API (MJ): 

First, this scenario requires subscriptions to third party tools. 

- Elevenlabs, for generating the voiceover (CLICK HERE) - they do have a free tier. - Imagineapi.dev for Midjourney API (CLICK HERE) - no free tier, their least expensive plan is $40 per month. 

- Midjourney to generate images (CLICK HERE) - at least the $10 a month plan. 

- Transloadit to create the slideshow video (CLICK HERE) - free tier is fine. - CloudConvert to convert files (CLICK HERE) - free tier is fine. - OpenAI for GPT…of course

Step one, create your Google Sheet (then connect it to this scenario): 

Step two, modify the GPT module as needed. Currently it works like this:

It corresponds with the Google Sheet. Text is the text that will be turned into an audio file (“read” by AI). Type of text is what the text is (a poem, a book chapter, etc). Representative is WHO is reading the text (the author, etc). Descriptor of info just tells the A.I. what they are reading (he describes a harrowing tale of a young elf…for example).

Step three, input your Imagineapi.dev API key in ALL of the Midjourney create and fetch modules. ALSO input your Imageinapi.dev URL (you get your own server URL upon signing up):

Step four, fill in your ElevenLabs API key AND the voice ID in the URL:

You’ll need to make a separate API call to retrieve all your voice IDs using these instructions here: CLICK HERE

If you have trouble, ElevenLabs support should be able to help. The voice ID is an alphanumeric string. 

Step five, in the value section under “params” in the Transloadit module, enter your API key after the word Bearer: 

ALSO, you’ll need to enter a Make.com webhook URL. The URL from the webhook in the second scenario for this the voiceover bots [simply duplicate] (see below). Please note, for this module, currently it creates a 10 image slideshow. You can modify this by adding or subtracting images from the code under “steps” as well as “use”. 

Step six, in the SECOND scenario for the voiceover bot, you’ll create a new webhook by clicking “Add” and then copy the URL it creates and paste it in the previous scenario where it says “MAKE.COM WEBHOOK”:

For the last module in THIS scenario, just make sure that the proper column is indicated…

Since the first module searches for rows that have no word ‘yes’ in them, this ensures the same info isn’t used more than once to generate a video.

Slideshow Video from Sheets Bot 2 (v2) 

This scenario takes the video from the webhook (that you generated in the first scenario) and transcribes the audio. 

*Alternatively you could save the text input from the Google Sheet to a Data Store and then just use THAT instead of using up tokens on transcription.* 

Finally, be sure to connect all of your social media accounts.** 

** If you want to publish to YouTube you will need a paid app from MSquare (CLICK HERE). The service is $20 per month OR a one-time lifetime subscription for $200.

You must fill out a form for the YouTube app, then it will be added to your Make account via API by MSquare. From there, you will receive detailed instructions on how to generate the Client ID and Secret needed to connect the app 

(CLICK HERE) 

When you have access to the app for YouTube, that will give you access to several modules, including “Upload a Video” (full list here: CLICK HERE). 

That should be all that is needed to generate faceless slideshow videos with a pre-written voiceover. 

Please note, this flow does not add subtitles. Transloadit DOES offer this service (and you would need to use Transloadit since they create the video in the first place….a third party service likely won’t work…we tried several), but it is only available on their $49 per month plan. 

Here are resources that may be helpful for this scenario or modifications you may wish to make: CLICK HERE

CLICK HERE

CLICK HERE

Slideshow Video from Sheets Bots using ModelsLab (ML) 

This scenario is IDENTICAL IN EVERY WAY to the Midjourney one, with one obvious exception. The API call is to Model Labs (formerly stablediffusionapi.com). This will require an account (CLICK HERE to sign up). It’s $30 per month for 13k images.

Note, in the ModelLabs/stablediffusionapi.com modules you need to input your API key…but you can also choose a different model id (this is the public model used to generate the images). 

Find them here: CLICK HERE

When you find one you like, click on it, and in the code snippet you’ll find the model_id.

You can also adjust other parameters, such as the size and number of samples (these are set to 4 because they use all 4 images from each prompt). 

More info about the dreambooth API here: 

CLICK HERE