Parsing

schema.parse(data);           // throws ZodError on failure
schema.safeParse(data);       // returns { success, data?, error? }
await schema.parseAsync(data);
await schema.safeParseAsync(data);

Type Inference

type MyType = z.infer<typeof mySchema>;    // output type
type MyInput = z.input<typeof mySchema>;   // input type
type MyOutput = z.output<typeof mySchema>; // same as z.infer

Primitives

z.string();
z.number();      // finite numbers only (no Infinity, no NaN)
z.bigint();
z.boolean();
z.symbol();
z.undefined();
z.null();
z.date();
z.nan();

Coercion

z.coerce.string();    // String(input)
z.coerce.number();    // Number(input)
z.coerce.boolean();   // Boolean(input)
z.coerce.bigint();    // BigInt(input)
z.coerce.date();      // new Date(input)

String Formats (Top-Level)

Use top-level functions, NOT methods. Methods like z.string().email() are deprecated.

// CORRECT (Zod 4)
z.email();
z.uuid();
z.url();
z.httpUrl();
z.hostname();
z.emoji();
z.base64();
z.base64url();
z.hex();
z.jwt();
z.nanoid();
z.cuid();
z.cuid2();
z.ulid();
z.ipv4();
z.ipv6();
z.mac();
z.cidrv4();
z.cidrv6();
z.hash("sha256");     // "sha1" | "sha384" | "sha512" | "md5"
z.e164();              // phone numbers
z.iso.date();
z.iso.time();
z.iso.datetime();
z.iso.duration();

// DEPRECATED (Zod 3 style — do NOT use)
z.string().email();    // ❌
z.string().uuid();     // ❌
z.string().url();      // ❌

UUID versions

z.uuid();                    // any RFC 9562/4122 UUID
z.uuid({ version: "v4" });  // specific version
z.uuidv4();                  // convenience
z.uuidv6();
z.uuidv7();
z.guid();                    // any 8-4-4-4-12 hex pattern (less strict)

Custom email regex

z.email();                                     // default (Gmail rules)
z.email({ pattern: z.regexes.html5Email });    // browser-style
z.email({ pattern: z.regexes.rfc5322Email });  // RFC 5322
z.email({ pattern: z.regexes.unicodeEmail });  // intl emails

JWTs

z.jwt();
z.jwt({ alg: "HS256" });

Numbers & Integers

z.number();      // any finite number
z.int();         // safe integer range
z.int32();       // int32 range
z.float32();     // float32 range
z.float64();     // float64 range

// bigint variants
z.bigint();
z.int64();
z.uint64();

Number validations

z.number().gt(5);
z.number().gte(5);          // alias: .min(5)
z.number().lt(5);
z.number().lte(5);          // alias: .max(5)
z.number().positive();
z.number().nonnegative();
z.number().negative();
z.number().nonpositive();
z.number().multipleOf(5);   // alias: .step(5)

String validations

z.string().max(5);
z.string().min(5);
z.string().length(5);
z.string().regex(/pattern/);
z.string().startsWith("abc");
z.string().endsWith("xyz");
z.string().includes("---");
z.string().uppercase();
z.string().lowercase();
z.string().trim();
z.string().toLowerCase();
z.string().toUpperCase();
z.string().normalize();

Objects

z.object({ name: z.string(), age: z.number() });          // strips unknown keys
z.strictObject({ name: z.string() });                      // errors on unknown keys
z.looseObject({ name: z.string() });                       // passes unknown keys through

Deprecated: .strict(), .passthrough(), .strip(), .merge(), .deepPartial().

Object methods

schema.extend({ newField: z.string() });    // add fields
schema.pick({ name: true });                // pick fields
schema.omit({ age: true });                 // omit fields
schema.partial();                           // all optional
schema.partial({ name: true });             // some optional
schema.required();                          // all required
schema.keyof();                             // ZodEnum from keys
schema.shape.name;                          // access inner schema

Prefer spread syntax for best tsc performance:

z.object({ ...Base.shape, ...Extra.shape, newField: z.string() });

Recursive objects

const Category = z.object({
  name: z.string(),
  get subcategories() {
    return z.array(Category);
  },
});

Enums

z.enum(["A", "B", "C"]);               // string enum
z.enum(MyTSEnum);                       // TypeScript enum (replaces z.nativeEnum())
z.enum({ A: 0, B: 1 } as const);       // enum-like object

Deprecated: z.nativeEnum(). Use z.enum() instead.

Arrays, Tuples, Sets, Maps

z.array(z.string());
z.array(z.string()).min(1).max(10).length(5);
z.tuple([z.string(), z.number()]);
z.tuple([z.string()], z.number());    // variadic: [string, ...number[]]
z.set(z.number());
z.map(z.string(), z.number());

Unions & Intersections

z.union([z.string(), z.number()]);
z.discriminatedUnion("status", [
  z.object({ status: z.literal("ok"), data: z.string() }),
  z.object({ status: z.literal("err"), error: z.string() }),
]);
z.xor([z.string(), z.number()]);          // exclusive union (exactly one match)
z.intersection(schemaA, schemaB);

Records

z.record(z.string(), z.number());                    // Record<string, number>
z.record(z.enum(["a", "b"]), z.string());            // exhaustive: { a: string; b: string }
z.partialRecord(z.enum(["a", "b"]), z.string());     // partial: { a?: string; b?: string }
z.looseRecord(z.string().regex(/^x_/), z.number());  // pass through non-matching keys

Breaking: z.record(z.string()) single-arg form is removed. Always pass both key and value schemas.

Literals

z.literal("hello");
z.literal(42);
z.literal(true);
z.literal(["a", "b", "c"]);   // multi-value literal (new in Zod 4)

Files

z.file();
z.file().min(10_000);                          // min size in bytes
z.file().max(1_000_000);                       // max size in bytes
z.file().mime("image/png");                    // single MIME
z.file().mime(["image/png", "image/jpeg"]);    // multiple MIMEs

Stringbool

z.stringbool();    // "true"/"yes"/"1"/"on"/"y"/"enabled" → true, inverses → false
z.stringbool({ truthy: ["yes", "y"], falsy: ["no", "n"] });
z.stringbool({ case: "sensitive" });

Template Literals

z.templateLiteral(["hello, ", z.string()]);                    // `hello, ${string}`
z.templateLiteral([z.number(), z.enum(["px", "em", "rem"])]);  // `${number}px` | ...

Optionals, Nullables, Defaults

z.optional(z.string());          // or z.string().optional()
z.nullable(z.string());          // or z.string().nullable()
z.nullish(z.string());           // optional + nullable

z.string().default("hello");     // short-circuits on undefined, returns output type
z.string().prefault("hello");    // pre-parse default, runs through validation
z.number().catch(42);            // fallback on any validation error

Breaking: .default() now expects output type, not input type. Use .prefault() for old behavior.

Transforms & Pipes

z.transform((val) => String(val));                      // standalone transform
z.string().transform((val) => val.length);              // pipe string → transform
z.preprocess((val) => String(val), z.string());         // transform → pipe to schema
z.string().pipe(z.transform((val) => val.length));      // explicit pipe

// .overwrite() — transform without changing inferred type
z.number().overwrite((val) => val ** 2);

Codecs (Bidirectional Transforms)

New in Zod 4.1. Define encode/decode pairs:

const stringToDate = z.codec(z.iso.datetime(), z.date(), {
  decode: (s) => new Date(s),
  encode: (d) => d.toISOString(),
});

stringToDate.decode("2024-01-15T10:30:00.000Z");  // → Date
stringToDate.encode(new Date());                    // → string
stringToDate.parse("2024-01-15T10:30:00.000Z");    // → Date (same as decode at runtime)

Error Customization

Use unified error param (replaces message, errorMap, invalid_type_error, required_error):

z.string().min(5, { error: "Too short" });
z.string({ error: (issue) => issue.input === undefined ? "Required" : "Not a string" });
z.string({ error: (issue) => {
  if (issue.code === "too_small") return `Min ${issue.minimum}`;
}});

Deprecated: message, errorMap, invalid_type_error, required_error.

Metadata & Registries

z.string().meta({ id: "email", title: "Email", description: "User email" });
z.string().describe("A description");      // shorthand for .meta({ description: ... })

// Custom registries
const myReg = z.registry<{ description: string }>();
z.string().register(myReg, { description: "..." });

JSON Schema

z.toJSONSchema(schema);                                        // Zod → JSON Schema
z.toJSONSchema(schema, { target: "draft-07" });                // specific draft
z.toJSONSchema(schema, { target: "openapi-3.0" });             // OpenAPI 3.0
z.fromJSONSchema(jsonSchema);                                  // JSON Schema → Zod (experimental)

Refinements

z.string().refine((val) => val.includes("@"), { error: "Must contain @" });
z.string().refine((val) => val.includes("@"), { error: "...", abort: true }); // stop on failure
z.array(z.string()).superRefine((val, ctx) => {
  ctx.addIssue({ code: "custom", message: "...", input: val });
});

Refinements now live inside schemas (not ZodEffects wrapper). You can interleave .refine() with other methods:

z.string().refine(v => v.includes("@")).min(5);  // works in Zod 4!

Error Pretty-Printing

z.prettifyError(zodError);   // formatted multi-line string
z.treeifyError(zodError);    // tree structure (replaces .format() and .flatten())

Further Reference

Zod Mini: See references/zod-mini.md for tree-shakable API differences, .check() usage, and bundle size tradeoffs
Codecs: See references/codecs.md for bidirectional transforms, encoding behavior, and common codec patterns (stringToDate, jsonCodec, etc.)
JSON Schema: See references/json-schema.md for z.toJSONSchema() options, format conversion, registry-based multi-schema, and z.fromJSONSchema()
Advanced patterns: See references/advanced.md for registries, refinement when, .superRefine(), .check(), functions, branded types, readonly, custom schemas, and advanced string/object/record/union options
Migration from Zod 3: See references/migration.md for all breaking changes and deprecated APIs

zod

Zod 4

Import

Core Patterns