Managed apps / AI assistants & agents
Managed Langflow_
Build AI agents on a visual canvas, call them by API
Langflow is a visual builder for AI agents and LLM workflows. Drag components onto a canvas, connect them, test in the playground and call the result through an API.
From £20 a month, VAT included. No server to run.
- HTTPS and your own domain
- Health checks and restarts
- Security updates underneath
- Prototyping agents
- Retrieval pipelines
- AI features behind an API
- Experimenting with prompts
First steps
- Sign in. When the app is live, open it in the console and choose Reveal admin password; it is shown once, so save it in your password manager. Then sign in at your-name.apps.fresh.dev/login. Username admin
- Add your model provider’s API key as a global variable.
- Start from a template, test it in the Playground, and create an API key under Settings.
- Activate users you create; they start inactive.
Run a flow by API
Every flow has an ID. Call it with your API key:
curl -X POST https://your-name.apps.fresh.dev/api/v1/run/YOUR_FLOW_ID \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"input_value": "Hello", "input_type": "chat", "output_type": "chat"}'Good to know
- Built-in code-running components are blocked, and only admins can add custom components.
The Langflow project
Langflow is open-source software, licensed under MIT. Fresh runs version 1.12.6 of the upstream release for you. Fresh is not affiliated with the Langflow project; for questions about the software itself, its documentation and community are the best place to start.
