How to convert JSON to Excel, nested JSON included

Pick a route by how deeply nested your records are and whether you write code. Excel can read a JSON file itself, rows.page flattens it in your browser and exports CSV, and pandas writes .xlsx from a script.

Excel has a JSON connector in Microsoft 365 on Windows and Mac. Power Query detects the table in the file and flattens what it can.

1. In Excel, choose Data > Get Data > From File > From JSON.
2. Pick the .json file and select Open. The Power Query Editor opens with the detected table.
3. If a column still shows Record or List, click the expand button in its header to pick fields or turn list items into rows.
4. Select Close & Load to put the table on a worksheet.

The connector doesn't read JSON Lines as is; Microsoft's workaround is M code with Lines.FromBinary and Json.Document. For JSONL files, or JSON with many nested levels, flatten the file in rows.page first.

rows.page opens .json, .jsonl, .ndjson and .ldjson files in your browser with DuckDB-WASM, or raw JSON you paste with Paste data. Nothing is uploaded unless you press Share.

1. Drop the file on rows.page. A .json file can be up to 1.5 GB per browser tab; a bigger JSONL file loads its first rows.
2. Check the columns: nested objects become dot-path columns such as user.address.city, and a wrapper like {"data": [...]} is unwrapped to its biggest list.
3. Hide fields you don't need in the Columns panel, or narrow the rows with the filter box.
4. Choose Export > CSV.
5. In Excel, choose Data > Get Data > From File > From Text/CSV, pick the CSV, set File Origin to 65001: Unicode (UTF-8) and select Load. Then use File > Save As to save the workbook as .xlsx.

The File Origin step matters because Excel only assumes UTF-8 when a CSV starts with a byte order mark (BOM), and rows.page exports UTF-8 without one. Double-clicking the CSV instead can turn é into é. Text cells that start with =, +, - or @ get a leading apostrophe so Excel doesn't run them as formulas.

Arrays inside a record stay as list columns: the grid shows the first 6 items and +N, and the CSV export writes each list as JSON text in one cell. For one row per array item, run SQL in the filter box, with double quotes around column names that contain dots:

SELECT
  id,
  "user.address.city" AS city,
  unnest(tags) AS tag
FROM data

Rows whose list is empty drop out of an unnest result. A SQL result shows and exports up to 1,000,000 rows, which fits in one Excel sheet.

An Excel sheet holds 1,048,576 rows by 16,384 columns, while the rows.page grid has no row limit. A filtered view exports every matching row, so narrow a big file below that limit first: filter with country:NL, price:>20, name:* (has a value) or -name:* (is empty), or click a histogram bar or top value in the Columns panel.

pandas writes .xlsx files directly through openpyxl (pip install pandas openpyxl), with no CSV step:

import json
import pandas as pd

with open("input.json", "r", encoding="utf-8") as f:
    records = json.load(f)

pd.json_normalize(records).to_excel("out.xlsx", index=False)

pd.json_normalize joins nested keys with dots, like user.address.city in rows.page; pass record_path when the records sit under a wrapper key. Lists are written as Python text such as ['a', 'b'], not JSON. to_excel raises an error when the DataFrame has more than 1,048,576 rows or 16,384 columns.