🎆 The cheapest & easiest way to use the best AI image models

Good morning AInauts,

Welcome to the new edition of your favorite newsletter. Today, we're not in the mood for news, but we'll show you practical hacks that we use every day.

The two tools in question actually come from the world of AI automation. But even beginners can do very cool things with them... We'll also try to keep it short and sweet at under 1,000 words. Feel free to give us feedback!

That's what it's all about today:

• 🎆 The cheapest & easiest way to use the best AI image models

• ✍️ How to use all language models simultaneously for the best answer

• 🤯 The development of AI in the last 2 years, in one video

Here we go!

🎆 The cheapest & easiest way to use the best AI image models

There's a lot going on in the world of AI image models at the moment.

While FLUX has dominated recently and replaced the old top dogs such as Midjourney and Stable Diffusion, another new model came around the corner just last week.

Under the code name red_panda, the creators Recraft AI have achieved a great success. It is a very powerful image model, especially when it comes to brand styles or vector graphics.

But it is hard to keep track of all the tools and what their best use cases are ...

• Midjourney

• FLUX

• Stable Diffusion

• Recraft

• Ideogram

And the most important question: who wants to pay for all these subscriptions?

Don't despair, the AInauts got you covered! We're here to help you see the big picture.

If you've been reading us for a while, you'll know that we're big fans of the Replicate platform.

We've been using Replicate for a long time to use various AI models in our automations via API - from text to image and more.

So Replicate is actually a tech platform. But the good thing about it is that you can use almost any model without paying a subscription!

You add your credit card and only pay for what you use. A picture with Recraft, for example, only costs 4 cents.

And here comes the kicker: Replicate has just released a new playground in beta!

This means that you can try out all image models via a simple user interface. And it is super easy to use!

1. Select model

2. Enter prompt

3. Select a few settings and off you go!

As I said, you have access to the most important models and can add even more, such as your own LoRAs etc.

So, you can use all models in a simple interface, with no ongoing subscription, no obligations, no speed limit, etc. And for very little money!

✍️ How to use all language models simultaneously for the best answer

Now that we have talked about images, you may also be wondering: "And what about language models?"

In this area, it's the same as with image models. New ones are coming out all the time: Claude, Gemini, o1, Grok, Llama, ...

Questions upon questions: How do I know which one is the best? Do I have to pay a subscription for all of them?

We love to test the models against each other. And then we simply take the best result. Of course, we also use another tool from the world of AI automation for this.

There is of course the free Chatbot Arena with limited features - but we have another tool that we can't do without: OpenRouter Chat! It works similar to Replicate, you deposit your credit card and pay based on usage.

The good thing here is that you can easily select all models and work with them in a chat window. The most exciting feature for us is that the models can compete against each other.

Here we select all flagship models in one chat:

• Claude 3.5 Sonnet

• o1-preview

• Gemini Pro 1.5

• Grok Beta

Everyone answers our request. We can also extend the answers if we need a more comprehensive result. And we can even add more models!

The whole thing is brilliant, especially for one-shot prompts. And as we said, without a subscription, without obligation and for a very low price.

🤯 The development of AI in the last 2 years, in one video

We want to end today's issue with a little inspiration. It's quite a simple video, but it got us thinking for a few days now.

It's incredible how quickly AI has developed around text, images and video in the last two years:

(The cat on the right is a video ...)

With the current models and tools around Runway, Kling, FLUX & Co. it's hard to imagine that the cat on the left was the measure of all things just two years ago ...

What kind of development is in store for us in the next 2 years ...?

We are simply thrilled every week and very grateful to share our passion with you.

Hopefully you feel the same way and have just as much fun and take something away for yourself.

Thank you for joining us!

Reto & Fabian from the AInauts

P.S.: Follow us on social media - that motivates us to keep going 😁!
Twitter, LinkedIn, Facebook, Insta, YouTube, TikTok

Your feedback is essential for us. We read EVERY comment and feedback, just respond to this email. Tell us what was (not) good and what is interesting for YOU.