Use jq to Format JSON in Vim on Ubuntu (Prettify & Minify)

Paste a blob of minified JSON into Vim and it’s one unreadable line. Instead of reaching for an online formatter, pipe the buffer through jq — one command reflows the whole file, and it works the same over SSH where a GUI tool isn’t an option.

---

Step 1: Install jq on Ubuntu

Install jq using the default package manager:

sudo apt update
sudo apt install jq

Verify the installation:

jq --version

You should see:

jq-1.6

---

Step 2: Test jq in the Terminal

You can quickly test jq by formatting a JSON file:

jq . data.json

This will pretty-print the JSON output in the terminal.

---

Step 3: Use jq Inside Vim

Open a JSON file in Vim and run:

:%!jq .

This command sends the entire file through jq and replaces it with formatted JSON.

What :%! actually does

:%!cmd isn’t a jq feature — it’s a Vim one, and it’s worth understanding because it works with any command. % is the whole-file range, and !cmd filters those lines through an external program, replacing them with its output. So :%!jq . pipes the buffer into jq . and swaps in the result. Two things that follow from that:

• Filter a selection, not the whole file: visually select some lines and run :'<,'>!jq ..
• Undo saves you when the input is invalid: if jq errors and mangles the buffer, a single u puts it back, because it’s just a normal edit.

The filter also doesn’t have to be . — any jq program works, so :%!jq '.items' reduces the buffer to just that field.

---

Step 4: Minify JSON in Vim

To minify the JSON into a single line:

:%!jq -c .

The -c option outputs compact JSON, which is useful for logs or API payloads.

---

Optional: Format JSON Files from the Command Line

You can also format JSON files directly from the terminal:

jq . data.json > formatted.json

To overwrite the original file safely:

jq . data.json | sponge data.json

Note: sponge is included in the moreutils package.

---

Conclusion

Using jq with Vim on Ubuntu is a simple and efficient way to format JSON without relying on heavy IDE features. It works consistently across local machines, servers, and SSH environments — and needs no plugin, since the filtering is built into Vim. It pairs naturally with a Vim setup you already use every day.