Use the free JSON to Java Classes below: your data never leaves your browser — all work happens on your device, and nothing is uploaded to any server.
From Raw JSON to Java Classes in One Step
An API returns a block of JSON. You need Java classes that match it, with fields, types, getters, and setters. Writing those classes by hand is slow and error-prone, especially when the response has nested objects and arrays. One missed field means a null at runtime. One wrong type means a crash when the parser meets a value it did not expect. This tool reads your JSON and writes the classes for you, so you can spend your time on the logic of your app instead of on boilerplate.
Paste the JSON into the box, set the name of the root class, and press Convert. The tool walks the whole document. Every object becomes a class. Every array of objects becomes a list of that class. Primitive values become the closest Java type: text becomes String, whole numbers become int, decimals become double, and true or false become boolean. The result is a set of classes you can copy straight into your project.
How the Types Are Chosen
Type inference is the heart of the tool. A value like 1024 is an int. A value like 87.5 is a double. A value like true is a boolean. A string is always a String, even if it looks like a number, because JSON does not distinguish between a numeric string and ordinary text, and guessing would be dangerous. If your API sends an id as “1024” in quotes, the tool keeps it a String, and that is the safe choice.
Arrays follow the same logic. The tool looks at the first non-null item in an array and uses its type for the whole list. An array of strings becomes List of String. An array of objects becomes a list of a new class named after the field, singularized, so an orders array produces an Order class. An empty array has no item to learn from, so it becomes a list of Object: the tool says so rather than inventing a type.
Null values get the same honest treatment: a field whose sample value is null becomes Object. Change it to the real type once you know what the API actually sends, since samples often show null for fields that are usually filled, like a discount code. Treat the generated code as a strong draft, then review the null and empty fields before you ship.
Nested Objects and Class Names
Real API responses are rarely flat. A user has an address, an order has line items, and a line item has a product. The tool creates one class per nested object, named after the field that holds it: the address object becomes an Address class, and the orders array becomes an Order class. If two fields would produce the same class name, the tool numbers the second one.
Field names are converted to Java style: zipCode stays zipCode, zip_code becomes zipCode, and ZIP code becomes zipCode as well. Getters and setters follow the standard pattern, so zipCode gets getZipCode and setZipCode. With Jackson or Gson, these classes work with the default settings in most cases.
A Practical Workflow
Start with a real response, not a tiny example: copy the full JSON from your API client or the browser network tab, including the nested parts. Then read the output once from top to bottom, checking the nested classes and the fields that came from nulls or empty arrays. Rename the root class to something meaningful, like ApiResponse, instead of leaving it as Root.
Next, add the annotations your project needs: JsonProperty with Jackson on fields whose names differ from the JSON keys, or nothing at all with Gson, where default naming usually works. If your API sometimes sends a number as a string, change that field to String and convert it in your own code.
Keep the original JSON next to the generated code while you test. Send the same response through your parser and confirm that every field fills in. Pay special attention to numbers: an id that fits in an int today may exceed it tomorrow, and a price that arrives as 49.99 needs a double. If a field sometimes arrives as a whole number and sometimes as a decimal, declare it double.
Common Mistakes to Avoid
The most common mistake is converting a sample that is too small. A sample with one order cannot show whether the orders array ever contains a different shape, and a sample with no address cannot generate an Address class. Always convert the largest, most complete response you have. The second mistake is trusting inferred types blindly: review the output the way you would review a colleague’s code.
The third mistake is forgetting that JSON keys can contain characters Java identifiers cannot. Keys with spaces or dashes are converted to camelCase here, but your parser still needs to map the original key to the new name, with Jackson’s JsonProperty or Gson’s SerializedName. Without that mapping, the field will stay null even though the class looks right.
Privacy and Limits
Everything runs in your browser. Your JSON is never uploaded, never stored, and never sent to any server. That makes the tool safe for responses with real user data, though you should still avoid pasting secrets like tokens or passwords. For extremely large files, convert a trimmed copy that still contains every shape you need.