Zod to Json Schema

Looking for the exact opposite? Check out json-schema-to-zod

Summary

Does what it says on the tin; converts Zod schemas into JSON schemas!

• Supports all relevant schema types, basic string, number and array length validations and string patterns.
• Resolves recursive and recurring schemas with internal $refs.
• Also able to target legacy Open API 3.0 specification for path schemas (3.1 supports regular Json Schema).

Sponsors

If you enjoy this package, consider dropping a couple of bucks on my GitHub Sponsors page <3

Build AI apps and workflows with Retool AI retool.com

Usage

Basic example

1import { z } from "zod";
2import { zodToJsonSchema } from "zod-to-json-schema";
3
4const mySchema = z
5  .object({
6    myString: z.string().min(5),
7    myUnion: z.union([z.number(), z.boolean()]),
8  })
9  .describe("My neat object schema");
10
11const jsonSchema = zodToJsonSchema(mySchema, "mySchema");

Expected output

1{
2  "$schema": "http://json-schema.org/draft-07/schema#",
3  "$ref": "#/definitions/mySchema",
4  "definitions": {
5    "mySchema": {
6      "description": "My neat object schema",
7      "type": "object",
8      "properties": {
9        "myString": {
10          "type": "string",
11          "minLength": 5
12        },
13        "myUnion": {
14          "type": ["number", "boolean"]
15        }
16      },
17      "additionalProperties": false,
18      "required": ["myString", "myUnion"]
19    }
20  }
21}

Options

Schema name

You can pass a string as the second parameter of the main zodToJsonSchema function. If you do, your schema will end up inside a definitions object property on the root and referenced from there. Alternatively, you can pass the name as the name property of the options object (see below).

Options object

Instead of the schema name (or nothing), you can pass an options object as the second parameter. The following options are available:

Option	Effect
name?: string	As described above.
nameStrategy?: "ref" | "title"	Adds name as "title" meta instead of as a ref as described above
basePath?: string[]	The base path of the root reference builder. Defaults to ["#"].
$refStrategy?: "root" | "relative" | "seen" | "none"	The reference builder strategy; "root" resolves $refs from the root up, ie: "#/definitions/mySchema". "relative" uses relative JSON pointers. See known issues! "seen" reuses the output of any "seen" Zod schema. In theory it's a more performant version of "none", but in practice this behaviour can cause issues with nested schemas and has now gotten its own option. "none" ignores referencing all together, creating a new schema branch even on "seen" schemas. Recursive references defaults to "any", ie {}. Defaults to "root".
effectStrategy?: "input" | "any"	The effects output strategy. Defaults to "input". See known issues!
dateStrategy?: "format:date" | "format:date-time" | "string" | "integer"	Date strategy, integer allow to specify in unix-time min and max values. "format:date" creates a string schema with format: "date". "format:date-time" creates a string schema with format: "date-time". "string" is intepreted as "format:date-time". "integer" creates an integer schema with format "unix-time" (unless target "openApi3" is used min max checks are also respected)
emailStrategy?: "format:email" | "format:idn-email" | "pattern:zod"	Choose how to handle the email string check. Defaults to "format:email".
base64Strategy?: "format:binary" | "contentEnconding:base64" | "pattern:zod"	Choose how to handle the base64 string check. Defaults to "contentEncoding:base64" as described here. Note that "format:binary" is not represented in the output type as it's not part of the JSON Schema spec and only intended to be used when targeting OpenAPI 3.0. Later versions of OpenAPI support contentEncoding.
definitionPath?: "definitions" | "$defs"	The name of the definitions property when name is passed. Defaults to "definitions".
target?: "jsonSchema7" | "jsonSchema2019-09" | "openApi3"	Which spec to target. Defaults to "jsonSchema7"
strictUnions?: boolean	Scrubs unions of any-like json schemas, like {} or true. Multiple zod types may result in these out of necessity, such as z.instanceof()
definitions?: Record<string, ZodSchema>	See separate section below
errorMessages?: boolean	Include custom error messages created via chained function checks for supported zod types. See section below
markdownDescription?: boolean	Copies the description meta to markdownDescription
patternStrategy?: "escape" | "preserve"	The Zod string validations .includes(), .startsWith(), and .endsWith() must be converted to regex to be compatible with JSON Schema's pattern. For safety, all non-alphanumeric characters are escaped by default (consider z.string().includes(".")), but this can occasionally cause problems with Unicode-flagged regex parsers. Use preserve to prevent this escaping behaviour and preserve the exact string written, even if it results in an inaccurate regex.
applyRegexFlags?: boolean	JSON Schema's pattern doesn't support RegExp flags, but Zod's z.string().regex() does. When this option is true (default false), a best-effort is made to transform regexes into a flag-independent form (e.g. /x/i => /[xX]/ ). Supported flags: i (basic Latin only), m, s.
pipeStrategy?: "all" | "input" | "output"	Decide which types should be included when using z.pipe, for example z.string().pipe(z.number()) would return both string and number by default, only string for "input" and only number for "output".
removeAdditionalStrategy?: "passthrough" | "strict"	Decide when additionalProperties should be false - whether according to strict or to passthrough. Since most parsers would retain properties given that additionalProperties = false while zod strips them, the default is to strip them unless passthrough is explicitly in the schema. On the other hand, it is useful to retain all fields unless strict is explicit in the schema which is the second option for the removeAdditional
override?: callback	See section

Definitions

The definitions option lets you manually add recurring schemas into definitions for cleaner outputs. It's fully compatible with named schemas, changed definitions path and base path. Here's a simple example:

1const myRecurringSchema = z.string();
2const myObjectSchema = z.object({ a: myRecurringSchema, b: myRecurringSchema });
3
4const myJsonSchema = zodToJsonSchema(myObjectSchema, {
5  definitions: { myRecurringSchema },
6});

Result

1{
2  "type": "object",
3  "properties": {
4    "a": {
5      "$ref": "#/definitions/myRecurringSchema"
6    },
7    "b": {
8      "$ref": "#/definitions/myRecurringSchema"
9    }
10  },
11  "definitions": {
12    "myRecurringSchema": {
13      "type": "string"
14    }
15  }
16}

Error Messages

This feature allows optionally including error messages created via chained function calls for supported zod types:

1// string schema with additional chained function call checks
2const EmailSchema = z.string().email("Invalid email").min(5, "Too short");
3
4const jsonSchema = zodToJsonSchema(EmailSchema, { errorMessages: true });

Result

1{
2  "$schema": "http://json-schema.org/draft-07/schema#",
3  "type": "string",
4  "format": "email",
5  "minLength": 5,
6  "errorMessage": {
7    "format": "Invalid email",
8    "minLength": "Too short"
9  }
10}

This allows for field specific, validation step specific error messages which can be useful for building forms and such. This format is accepted by react-hook-form's ajv resolver (and therefor ajv-errors which it uses under the hood). Note that if using AJV with this format will require enabling ajv-errors as vanilla AJV does not accept this format by default.

Custom Error Message Support

• ZodString
  • regex
  • min, max
  • email, cuid, uuid, url
  • endsWith, startsWith
• ZodNumber
  • min, max, lt, lte, gt, gte,
  • int
  • multipleOf
• ZodSet
  • min, max
• ZodArray
  • min, max

override option

This options takes a Zod schema definition, the current reference object (containing the current ref path and other options), an argument containing inforation about wether or not the schema has been encountered before, and a forceResolution argument.

Since undefined is a valid option to return, if you don't want to override the current item you have to return the ignoreOverride symbol exported from the index.

1import zodToJsonSchema, { ignoreOverride } from "zod-to-json-schema";
2
3zodToJsonSchema(
4  z.object({
5    ignoreThis: z.string(),
6    overrideThis: z.string(),
7    removeThis: z.string(),
8  }),
9  {
10    override: (def, refs) => {
11      const path = refs.currentPath.join("/");
12
13      if (path === "#/properties/overrideThis") {
14        return {
15          type: "integer",
16        };
17      }
18
19      if (path === "#/properties/removeThis") {
20        return undefined;
21      }
22
23      // Important! Do not return `undefined` or void unless you want to remove the property from the resulting schema completely.
24      return ignoreOverride;
25    },
26  },
27);

Expected output:

1{
2  "type": "object",
3  "required": ["ignoreThis", "overrideThis"],
4  "properties": {
5    "ignoreThis": {
6      "type": "string"
7    },
8    "overrideThis": {
9      "type": "integer"
10    }
11  },
12  "additionalProperties": false
13}

Known issues

• When using .transform, the return type is inferred from the supplied function. In other words, there is no schema for the return type, and there is no way to convert it in runtime. Currently the JSON schema will therefore reflect the input side of the Zod schema and not necessarily the output (the latter aka. z.infer). If this causes problems with your schema, consider using the effectStrategy "any", which will allow any type of output.
• JSON Schemas does not support any other key type than strings for objects. When using z.record with any other key type, this will be ignored. An exception to this rule is z.enum as is supported since 3.11.3
• Relative JSON pointers, while published alongside JSON schema draft 2020-12, is not technically a part of it. Currently, most resolvers do not handle them at all.
• Since v3, the Object parser uses .isOptional() to check if a property should be included in required or not. This has the potentially dangerous behavior of calling .safeParse with undefined. To work around this, make sure your preprocess and other effects callbacks are pure and not liable to throw errors. An issue has been logged in the Zod repo and can be tracked here.

Versioning

This package does not follow semantic versioning. The major and minor versions of this package instead reflects feature parity with the Zod package.

I will do my best to keep API-breaking changes to an absolute minimum, but new features may appear as "patches", such as introducing the options pattern in 3.9.1.

Changelog

https://github.com/StefanTerdell/zod-to-json-schema/blob/master/changelog.md