Last verified
Data Analyst GPT
Editor's PickUpload CSV, Excel, or JSON and get charts, statistics, and summaries without writing code
Conversation starters
- Upload a CSV and summarize the key statistics
- Which columns have the most missing values?
- Create a bar chart of sales by region
- Find the correlation between price and quantity
"Exploratory data analysis in 30 seconds: upload, ask, get a chart."
What is the Data Analyst GPT?
The Data Analyst GPT is a first-party GPT built by OpenAI that runs Code Interpreter in a data and databases analysis context. Upload a CSV, Excel file, or JSON, ask a question in plain English, and it writes Python, executes it in a sandboxed environment, and shows the result as a summary, table, or chart.
How does it work?
When you upload a file, the GPT loads it into a sandboxed Python environment with pandas. Ask a question in natural language, such as “which product category has the highest return rate?”, and it translates the question into Python code, runs it, and presents the output. Follow-up questions let you change chart styles, filter the data differently, or request statistical tests, all without writing code yourself.
For analysis that needs to connect findings to published research, pair the Data Analyst GPT with Consensus: pull the relevant literature first, then examine your data against what the studies show.
When should you use it?
Use the Data Analyst GPT for one-off exploratory analysis of structured data files when you need a chart or summary quickly and don’t want to open a Jupyter notebook. It handles datasets up to tens of thousands of rows well; for larger datasets, pre-aggregate before uploading.
Frequently asked questions
What file formats does the Data Analyst GPT accept?
It accepts CSV, Excel (.xlsx and .xls), JSON, plain text, and PDF files. For very large files over 50MB you may need to split the data into smaller chunks. The Code Interpreter runs in a sandboxed Python environment with pandas, matplotlib, and seaborn available.
Can the Data Analyst GPT connect to live databases?
No. The base GPT works only with uploaded files. For live database access you need a custom GPT with database actions configured, or export a query result to CSV and upload it. Because nothing leaves the uploaded file, it's a reasonable option for sensitive data.
Does the Data Analyst GPT save my uploaded data between sessions?
Files are available for the duration of a conversation but are not persisted across sessions. Each new conversation starts with a clean Python environment. If you need to reanalyze the same dataset, re-upload it at the start of the session.