This month, we bring you better navigation, a simpler getting-started experience, and several smaller quality-of-life improvements. Without further ado, here’s what’s new.
Revamped AI Studio Library
The agent library got a proper refresh with clearer categories and easier browsing. Every workflow now surfaces more metadata at a glance, including tags, categories, and how popular it is with other teams, so you can tell what’s worth trying before you open it.
The AI Assistant is also back and better than before. It sits right at the top of the library for quick access. Simply describe the agent you need in plain language, and it automatically builds a working version for you. From there, all you have to do is adjust it to your liking and connect your own tools and integrations.
New AI Projects Library
AI Projects also got its own library of pre-made workflows. It works much like the AI Studio library, but with a quicklink straight to the AI Projects setup wizard. There you can describe the project you want, and get a full team of agents built automatically.
Hire an entire team from the library, and you’re up and running in no time. We’re adding more regularly, so there’s always something new to try.
Ready to grow your business?
Start your free trial today and see results within days.
MCP & Custom LLM Improvements
Bringing your own model (Pro) or connecting to external MCP servers is simpler now. Both configure directly in your workspace settings, and once set up, they’re available to use in AI Projects too, not just regular flows.
For custom models, you can now choose from:
- OpenAI — connect your own API key to use OpenAI’s models under your own account and rate limits.
- Ollama — point FlowHunt at a self-hosted Ollama instance to run open models on your own infrastructure.
- NVIDIA NIM — use NVIDIA’s optimized inference microservices for your custom or open-weight models.
- OpenRouter — route to any model available through OpenRouter’s aggregated catalog from a single connection.
Better Credit Management
We’ve added more control over how credits are managed across your workspaces:
- Workspace limits — set a spending cap per workspace, so no single project can run away with the whole team’s budget.
- Low-balance alerts — get notified before a workspace runs out of credits, so you can keep your workflows running.
- Auto-recharge — top up credits automatically when a workspace dips below a threshold you set, to ensure your workflows never go dark.
And that’s a wrap for May. We continue pushing AI Projects forward, with some bigger things just around the corner, so stay tuned!
