Since launching the AI Agent Builder last month, we’ve been hard at work adding features—and we’ve loved seeing your feedback along the way. We wanted to share some of our latest improvements designed to make the AI Agent Builder more functional for you.
But before we jump into updates, here’s a fun fact: the most-forked collection is Customer Service Agent, which dynamically responds to queries, routes decisions, and processes responses. If you haven’t checked it out yet, now’s a great time! Check it out.
Now, on to what’s new:
DeepSeek and OpenAI Evaluation – We’ve added support for DeepSeek (deepseek-reasoner, deepseek-chat) and the latest OpenAI (o1, o1-mini, o3-mini) models, letting you compare responses, cost, and performance to pick the best one for your use case.
Python Support in ToolGen – You can now generate integration-ready Python code for the Agent’s tools.
And there’s more on the way:
- Cloud Deployment – Run your agents in the cloud and integrate seamlessly with external services.
- Monitoring – Get full visibility into your deployed agents to ensure optimal performance.
Want to be the first to hear about upcoming features? If you haven’t already, subscribe for email updates.
As always, keep the feedback coming—we’re building this with you!