Start here: run your first agent

LESSONLesson 0 · ~20 min

🎯Goal. Install Langflow, open the Simple Agent template, and watch an agent reason and pick a tool in the Playground.

▶ Try this prompt

I want to add 4 and 4.

Open Langflow → New Flow → pick the Simple Agent template → paste an OpenAI API key → click Playground and type the line above.

1Install Langflow free — the easiest paths are the Desktop app or Docker; you can also use pip/uv or run from source. It needs Python 3.10–3.14 and about 2GB RAM minimum.
2Open Langflow, click New Flow, and choose the Simple Agent template. Paste an OpenAI API key into the model block (you bring your own key — Langflow itself is free).
3Click Playground, type "I want to add 4 and 4." and send. The agent shows its reasoning, picks the Calculator tool, and answers 8 — you watch it think step by step.

A Langflow Agent component shown as editable Python (class AgentComponent) on the left and the Agent node's configuration panel on the right with Role, Language Model, Tools and Input Message fields — Every block is real, editable Python (left) behind a simple config panel (right) — the Agent node exposes **Role**, **Language Model**, **Tools** and **Input Message** without making you touch code. Source: https://www.langflow.org

✓You'll see. A running agent that reasons out loud, chooses the Calculator tool, and returns 8 — assembled from a template without writing code.

💳Cost. Langflow is free and open-source (MIT) — you self-host it at no cost. You pay only your own infrastructure plus the LLM API key you bring (here, a few cents of OpenAI usage).

💡Takeaway. Langflow turns a template into a working, tool-using agent in minutes — free to run, your key, your machine.