Execute Python scripts

[rprince764]

Hi team, Is there any way I can run python scripts in the flowise workflow?

Say this is my scenario:

1. Run the python script
2. Send the output to AI/LLM
3. Capture the output of AI/LLM
4. Display to user

[OMEE-69]

How to Run Python Scripts in Flowise Workflow

Scenario:

I want to:

1. Run a Python script

2. Send the script output to an LLM

3. Capture the LLM's output

4. Display the result to the user

---

Solution Options

🔹 Option 1: Use Custom Function Node + External Python API

Recommended if you're hosting Python scripts or have a backend server.

Steps:

1. Create an API using Flask or FastAPI to run your Python script and return the output.

2. In Flowise, use the “API” node or “Custom Function” node to call this API.

3. Send the API response to the LLM Node (OpenAI, Anthropic, etc.).

4. Display the final response to the user using a UI node or Chat Interface.

Visual Flow:

[User Input]
↓
[API Node] → calls your Python script
↓
[LLM Node] → Receives output from script
↓
[Response Node] → Displays final output

Pros:

Simple and modular

Supports any Python logic

Scalable

---

🔹 Option 2: Use Langchain's PythonREPLTool in a Custom Agent

Flowise uses Langchain under the hood. If you’re self-hosting, you can create a custom agent that uses Langchain's PythonREPLTool.

How:

Create a custom node or agent

Enable Python code execution inside it

Send output to LLM and show result

Requirements:

Self-hosted Flowise

Intermediate Langchain knowledge

---

🔹 Option 3: Use Node.js child_process to Execute Python

If you're customizing Flowise nodes (Node.js environment), you can use:

const { exec } = require('child_process');

exec('python3 script.py', (error, stdout, stderr) => {
if (error) {
console.error(exec error: ${error});
return;
}
console.log(Python output: ${stdout});
});

Then return the output back to Flowise nodes.

---

Summary Table

Task Tool / Method

Run Python script Flask / FastAPI endpoint
Connect to Flowise API Node / Custom Function
Process with LLM OpenAI / Claude / etc.
Show result to user Chat Interface / UI Node

Python Script with FastAPI

Create a FastAPI server to run your Python code.

main.py

from fastapi import FastAPI
from pydantic import BaseModel
import subprocess

app = FastAPI()

class InputData(BaseModel):
command: str # Python command to run

@app.post("/run-script")
def run_script(input_data: InputData):
try:
result = subprocess.run(
["python3", "-c", input_data.command],
capture_output=True,
text=True
)
return {"output": result.stdout.strip(), "error": result.stderr.strip()}
except Exception as e:
return {"error": str(e)}

** Install Dependencies**

pip install fastapi uvicorn

▶️ Run the Server

uvicorn main:app --host 0.0.0.0 --port 8000

[rprince764]

This is awesome!! Thanks a lot!