# ChatGPT Code Interpreter vs Tableau | ChatGPT, Tableau GPT & LLMs

> This is content from just-tim, the data-and-analytics channel by Tim Ngwena (formerly 'Tableau Tim'). Tim has 12+ years of hands-on BI experience and covers Tableau most of all, plus Power BI, Looker, Hex, SQL and data modelling, the analytics industry, and the craft of doing the job — always tool-agnostic and honest about the trade-offs.

- **Author:** Tim Ngwena (just-tim, https://just-tim.com/about)
- **Published:** 2023-07-11
- **Format:** Video · 1472 min watch · transcript available
- **Topics:** AI & ML, Data visualisation, Tool strategy
- **Tools:** Python (pandas); Tableau
- **Canonical:** https://just-tim.com/posts/chatgpt-code-interpreter-vs-tableau-chatgpt-tableau-gpt-llms
- **Watch:** https://www.youtube.com/watch?v=4P5VCHg-KAU

Over the weekend OpenAI added the Code Interpreter to ChatGPT, a separate GPT-4 model that runs Python behind the scenes to do data analysis. I put it head to head with Tableau using the Superstore sample data to see how close it gets to the numbers and charts, and whether the analysis is reliable.

## Key takeaways

- Code Interpreter is a separate GPT-4 model you enable in Settings under the beta section, and it can't be used at the same time as plugins.
- It works by uploading your file and writing Python (pandas) against it, so you should never upload sensitive or untrusted work data.
- It nailed total sales, profit ratio, average discount and a tightly filtered query (Bavaria, central region, specific dates) to exact Tableau figures, though it defaulted to euros with no basis.
- It can render styled HTML pivot tables and seaborn scatterplots, but struggled to reproduce nested rows/column sections exactly and couldn't draw maps because it has no internet access.
- The real threat to Tableau and Power BI isn't the model itself but the text box interface, which lets anyone who asks good questions do analysis without drag and drop.

## Chapters

- 0:00 What Code Interpreter is
- 0:28 Enabling the feature and plugins
- 1:11 Uploading data and first look
- 4:22 Total sales, profit ratio and discount
- 8:29 A complex filtered query
- 12:28 Why the text box is the real threat
- 14:53 Building tables and scatterplots
- 21:48 Map limitations and internet firewall
- 23:27 Closing thoughts

Watch the full video, read the transcript and use chapter deep-links on the page: https://just-tim.com/posts/chatgpt-code-interpreter-vs-tableau-chatgpt-tableau-gpt-llms

---
just-tim — Data and analytics, with a point of view. · https://www.youtube.com/channel/UC7HYxRWmaNlJux-X7rNLZyw · https://twitter.com/TableauTim · https://www.linkedin.com/in/timngwena