Hey friend,
Bettina here. It's the middle of summer and it’s hot, so it’s firmly iced coffee weather. The excessive consumption of cold caffeinated drinks is probably questionable, but it’s fueling some of the truly fun stuff we’ve been shipping that I wanted to tell you about: an out of the box way to help your users create content with AI, a look at how we got our own AI agent's token bills under control, and sample code worth stealing for multi-language builds.
Let’s dive right in!
An email design agent you can launch today (Beta)
Changelog highlight
If you'd like to help your users create content faster with the help of AI but don't want (or don’t have the time) to build an agent from scratch, you might enjoy our latest update.
The Beefree AI AddOn is a ready-to-use AI copilot that can create designs from scratch, edit layouts, generate variations, or support with copywriting. You enable the AddOn in your Developer Console, connect it to your favorite AI provider (we currently support Anthropic, Gemini, and OpenAI), and done. Your users can now create and update their content with simple prompts.
The Beefree AI AddOn is now available in Beta. Give it a try and let us know what you think.
Multi-language templates
Sample code of the month
If your users design emails for more than one market, you've probably heard about this challenge: How can you create multiple (language) versions of the same content, without creating a whole separate email for each?
Our multi-language template example shows you how. Using the @beefree.io/react-email-builder as a base, it gives you an email builder that makes it easy to swap between 10 different languages, including right-to-left (RTL) languages.
And if you want to take the actual translation work off your users plate as well, pair multi-language templates with the DeepL AddOn and your users get automatic translations that work seamlessly with multi-language templates.
Building AI agents for email creation: Best practices for saving tokens
From our reading list, to yours
Our team spent months building an email creation agent on top of Beefree SDK's MCP Server, and while the results were great, the token bills, at first, were not (apologies again to our Head of Controlling 😅).
In this post, Marco and Marco* from our development team share how they reduced costs by up to 94% without sacrificing output quality, so you don't have to learn the same lessons. Their key tricks: collapsing 33 individual tool schemas to cut token use 68-96%, prompt caching (which shaved off another 30%), and working with template skeletons. If you’re building your own content creation agents, check out their learnings before you chalk up a big AI bill to "it is what it is".
What’s keeping you on your toes this summer? If you’re also thinking about how to use AI to supercharge your users’ content creation workflow, I’d love to hear more. Just hit reply!
Signing off,
Bettina Specht
Marketing, Beefree SDK
* In case you're wondering: no, not all our developers are called Marco. On our AI team though, the Marco density is high.
