YAML to JSON converter

How the conversion works

This direction is the harder one. YAML has features JSON simply does not have, so the converter has to decide what to do with them rather than copy them across.

Comments being dropped is the loss that surprises people most. If your YAML is documentation as much as configuration — and a well-written values.yaml usually is — round-tripping through JSON strips that documentation permanently. Keep the YAML as the source of truth and treat the JSON as a build artifact.

Anchors, expanded

Aliases are the clearest illustration of what "lossy" means here. The YAML on the left defines a block once and reuses it. The JSON on the right has no way to say "the same as above", so the block is repeated.

defaults: &defaults
  retries: 3
  timeout: 30

staging:
  <<: *defaults
  host: stg.example.com

production:
  <<: *defaults
  host: example.com

{
  "defaults": { "retries": 3, "timeout": 30 },
  "staging": {
    "retries": 3,
    "timeout": 30,
    "host": "stg.example.com"
  },
  "production": {
    "retries": 3,
    "timeout": 30,
    "host": "example.com"
  }
}

The JSON is correct and equivalent. It is also three times longer and no longer carries the intent that staging and production share a base. Converting back will not restore the anchors.

Doing the same thing in code

Python

import json, yaml

with open("config.yaml") as f:
    data = yaml.safe_load(f)          # safe_load, never load()

print(json.dumps(data, indent=2))

Use safe_load. Plain yaml.load() can construct arbitrary Python objects from a crafted document, which is a remote code execution path if the YAML came from anywhere you do not control.

Node.js

import fs from "node:fs";
import yaml from "js-yaml";

const data = yaml.load(fs.readFileSync("config.yaml", "utf8"));
console.log(JSON.stringify(data, null, 2));

Multiple documents in one file

# Python — a list of documents
docs = list(yaml.safe_load_all(open("manifests.yaml")))
print(json.dumps(docs, indent=2))

# Command line
yq -o=json '.' manifests.yaml

Errors you are likely to hit

Tabs used for indentation

YAML forbids tabs as indentation, full stop. An editor configured to insert tabs will produce a file that looks perfectly aligned and fails to parse. Set your editor to two spaces for .yml and .yaml.

Missing space after the colon

key:value is a single scalar string, not a mapping. YAML needs key: value. This one is easy to miss because most languages accept the compact form.

The value looks like something else

An unquoted 1.10 becomes the float 1.1, losing the trailing zero. An unquoted 3:00 can parse as a sexagesimal number under YAML 1.1. Version strings, times, and anything with a leading zero want quotes.

A colon inside an unquoted value

message: Error: not found is ambiguous and most parsers reject it. Quote the whole value: message: "Error: not found".

If you are staring at a parser error right now, the YAML error reference lists the common messages verbatim with the fix for each.

Nothing leaves this tab

The YAML people paste here is usually a values.yaml, a CI pipeline definition, or an Ansible playbook — files that sit in private repositories and name every host, queue, and service account an organisation runs. Even with the secrets templated out, the structure alone maps your infrastructure.

None of it reaches us. Parsing and conversion happen locally, in JavaScript, on your machine. We could not produce your document if someone asked us to, because we never received it.

Questions people ask

Related tools

• JSON to YAMLConvert JSON into readable block YAML
• YAML validatorCheck YAML syntax and see how it really parsed
• Common YAML errorsParser messages with the fix for each
• YAML formatterNormalise indentation and quoting in YAML
• JSON vs YAMLHow the two formats differ and when to use each
• JSON to TOMLConvert JSON to TOML config format
• Kubernetes JSON to YAMLConvert a manifest and check it before you commit it