JSON to Kotlin Data Class Generator

JSON to Kotlin: data classes from a sample

Kotlin's data class is the natural shape for JSON models: immutable properties, generated equals, hashCode, toString and copy. This JSON to Kotlin generator writes data classes from a sample, one per object shape, using val properties, Kotlin types, and nullable types where the data shows a value can be missing or null.

Example

{
  "id": 42,
  "title": "Write docs",
  "done": false,
  "priority": 2.5,
  "tags": ["docs"],
  "owner": { "name": "Ada", "email": null },
  "due": "2026-10-01T09:00:00Z",
  "subtasks": [
    { "title": "Outline", "estimate": 2 },
    { "title": "Draft" }
  ]
}

With Types only checked:

package quicktype

import java.time.OffsetDateTime

data class Task (
    val id: Long,
    val title: String,
    val done: Boolean,
    val priority: Double,
    val tags: List<String>,
    val owner: Owner,
    val due: OffsetDateTime,
    val subtasks: List<Subtask>
)

data class Owner (
    val name: String,
    val email: Any? = null
)

data class Subtask (
    val title: String,
    val estimate: Long? = null
)

Optional properties are nullable and default to null, so constructing a Subtask without an estimate compiles. The date string became java.time.OffsetDateTime, which is available on the JVM and on Android API 26+.

Jackson helpers in full mode

With Types only off, the output targets Jackson with the Kotlin module: properties get @get:JsonProperty(required=true) and @field:JsonProperty(required=true) annotations, a preconfigured mapper handles OffsetDateTime and skips nulls when writing, and each top-level class gets toJson() and Task.fromJson(jsonString). Add com.fasterxml.jackson.module:jackson-module-kotlin to your dependencies, as the header comment explains.

Using kotlinx.serialization, Moshi or Gson instead

The types-only classes are a good base for other libraries:

• kotlinx.serialization: annotate each class with @Serializable, add @SerialName("user_name") where keys differ, and use a String or a custom serializer for dates, since OffsetDateTime has no built-in serializer.
• Moshi: add @JsonClass(generateAdapter = true) and @Json(name = "..."), the usual choice in Android apps using Retrofit.
• Retrofit: declare suspend fun getTask(@Path("id") id: Long): Task in your service interface and register a converter factory for the library you chose; the data class is used as the return type directly.
• Gson: works with data classes, but it ignores Kotlin's null-safety and default values, so a missing non-null field can end up null at runtime.

Whatever library you use, property names are converted to camelCase (is_active becomes isActive), so add name annotations when the JSON uses snake_case.

Null safety and missing keys

Kotlin's type system makes a promise: a String is never null. JSON libraries must keep that promise when data is incomplete. With Jackson's Kotlin module, a missing value for a non-null property without a default causes a MismatchedInputException (reported for Kotlin as a missing parameter). kotlinx.serialization throws MissingFieldException. Both are good: the error appears at the boundary instead of as a NullPointerException later. If an API field is sometimes absent, make the property nullable with a default, like the generated estimate: Long? = null.

Type mapping

Replace the quicktype package with your own. For XML sources, use XML to Kotlin; for Java models, JSON to Java. Check the input with the JSON Validator. Generation runs in your browser.