n8n-code-python

Python Code nodes in n8n run Python using only the standard library. Write your transformation logic using built-in modules like json, datetime, re, base64, and urllib.parse, then return results in the required format: a list of item objects with a json property containing your transformed data.

Webhook data in Python Code nodes is accessed via `_json["body"]`, not at the root of `_json`. Use this path to safely read incoming webhook payloads, then transform and return them in the standardized list format n8n expects.

Python Code nodes must return a list of item dictionaries in the form `[{"json": {...}}]`. Each item's json property holds your transformed data. This structured format ensures downstream nodes receive data in the expected schema.

Use `_input.all()` for batch processing when you need to transform multiple items together. Use `_input.first()` for single-item flows. Choose Run Each Item mode when your logic must execute once per input item rather than across the entire batch.

Top errors include ModuleNotFoundError (using external libraries), KeyError (accessing missing dictionary keys), IndexError (out-of-bounds list access), Empty code or Missing return, and incorrect return format. Prevent them by using only stdlib modules, validating data before access, and always returning the required list-of-dicts structure.

Python (Beta) and Python (Native) are different runtime options in n8n with distinct behavior and library support. Choose based on your workflow requirements and n8n's documentation for which runtime best suits your data processing, transformation, and webhook handling tasks.