> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://humanloop.com/docs/v5/reference/python-environment/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://humanloop.com/_mcp/server.

> This reference provides details about the Python environment and supported packages.

Humanloop allows you to specify the runtime for your code [Evaluators](../explanation/evaluators) and [Tool](../explanation/tools) implementations in order
to run them natively with your Prompts in our Editor and UI based Evaluation workflows.

## Environment details

Python version: **3.11.4**

```
anthropic==0.29.0
continuous-eval==0.3.13
jellyfish==1.1.0
jsonschema==4.22.0
langdetect==1.0.9
nltk==3.8.1
numpy==1.26.4
openai==1.35.10
pandas==2.2.2
pydantic==2.8.2
requests==2.32.3
scikit-learn==1.5.1
spacy==3.7.5
sqlglot==25.5.1
syllapy==0.7.2
textstat==0.7.3
transformers==4.43.4
```

If you have any specific packages you would like to see here, please let us know at [support@humanloop.com](mailto:support@humanloop.com).

> Humanloop provides a secure Python runtime to support defining code based Evaluator and Tool implementations.