Startup Diaries #13: Scade community, audio summarizer and ChatGPT-4o-mini

Hey there, welcome back to another edition of the Scade.pro digest! We help you convert AI into revenue, engagement, and new projects. With over 1,500 models at your fingertips, you can test and integrate them, and speed up your time to market like never before.

We’ve got some cool stuff this week: community forum, a multimodal template with a little tutorial on how to modify it and save money with GPT-4o-mini.

Audio Summarizer Template

Today, let’s dive into how to use an audio summarizer template and why it’s going to be your new favorite one for daily tasks.

For example:

  • You can summarize meetings with your team and add them to your notetaker or a follow-up email.
  • Or after a webinar for your sales team, upload the recording and make a summary that you can add to your knowledge base.

How it works

  1. Upload your audio file

  2. Run the flow

  3. Convert audio to text using Whisper

  4. Summarize the text with ChatGPT

How you can modify it

You can add an extra field in the Start Node to type in a question. Then, modify the prompt in the ChatGPT node so the AI responds to whatever question you ask. Use a Prompt Generator to write the perfect prompt for this task — explain what you need and copy the final prompt into the ChatGPT node in Summarizer. And don’t forget to refer to your question using the expression editor.

This modified audio summarizer can turn into a content generator for various materials:

  • Take the webinar, we already discussed a bit earlier and create training materials and FAQs for onboarding new employees.
  • Upload an educational video and ask the template to draft a blog post for your corporate LinkedIn. By the way, one of our users created a special tool for that called Content Genie. We featured it in the #12th edition if you’re curious.

Oh, and pro tip: save some money by switching to the brand new GPT-4o-mini model. More details on that coming right up!

New Model: ChatGPT-4o-mini

This new model is over 60% cheaper than its older sibling, GPT-3.5 Turbo, and it totally outshines it in multimodality. It supports text, vision inputs, and outputs with plans to add audio and video in the future. It can handle up to 128,000 tokens — perfect for large documents or extended chats.

Some teams that create AI products with Scade have already switched to GPT-4o-mini and are loving the cost savings without sacrificing the quality. So, it’s the perfect time for you to start building your own templates with it.

Add this model to Compare LLMs or Martian LLM Router templates and see its performance by yourself.

By the way, we’ve added the option to use your own credentials in ChatGPT and Claude nodes. Or continue using ours as usual :slight_smile:

Scade Community

You know us, people behind Scade.pro. We’re probably LinkedIn, Product Hunt, or Twitter friends, right?

We decided to create a space where we can all hang out and share stories about using AI with Scade. It’s called Scade Forum and you can find it at community.scade.pro.

Join chats about Scade’s features, send your questions, and drop your feedback. If you came across a problem or solved a tricky issue — seek help or share your solutions.

Meet people who are just as passionate about tech and AI as you are. Who knows, you might find your new best friend here.

Can’t wait to chat with you all, because let’s be honest, who doesn’t love hours of talk about AI? :slight_smile:

Join the Forum

1 Like