Expressive
Over 150 built-in validators across strings, numbers, dates, arrays, objects, binaries, and more — with chainable rules that read like English.
joi
The most powerful schema description language and data validator for JavaScript
Skip to content
1. Schema Variants with
Get started in seconds
console
npm install joiconsole
yarn add joiconsole
pnpm add joijs
import Joi from 'joi';
const schema = Joi.object({
username: Joi.string().alphanum().min(3).max(30).required(),
email: Joi.string().email().required(),
age: Joi.number().integer().min(18),
});
const { error, value } = schema.validate({
username: 'danilo',
email: 'danilo@example.com',
age: 28,
});Built for real-world validation
1. Schema Variants with .alter() + .tailor()
Define one schema. Produce specialized versions for different contexts — like separate validation for GET vs POST — without duplicating anything.
js
const schema = Joi.object({
id: Joi.number().alter({
create: (s) => s.forbidden(),
update: (s) => s.required(),
}),
name: Joi.string().required(),
});
const createSchema = schema.tailor('create');
const updateSchema = schema.tailor('update');One source of truth. No drift between schemas.
2. Declarative Field Dependencies
Express relationships between fields without writing validation logic. Joi handles the permutations.
js
const schema = Joi.object({
email: Joi.string().email(),
phone: Joi.string().pattern(/^\+?[0-9]{7,15}$/),
address: Joi.string(),
})
.or('email', 'phone') // at least one contact method
.with('phone', 'address') // phone requires address
.xor('email', 'phone'); // but not bothSeven dependency methods: .with(), .without(), .or(), .and(), .xor(), .oxor(), .nand().
3. Cross-Field References
Reference other fields directly in validation rules. No callbacks, no workarounds.
js
const schema = Joi.object({
startDate: Joi.date().required(),
endDate: Joi.date().greater(Joi.ref('startDate')).required(),
password: Joi.string().min(8).required(),
confirm: Joi.any().valid(Joi.ref('password')).required(),
});References support relative paths, value transforms, and even context variables.
4. Rich, Actionable Errors
Every validation failure includes the exact path, a machine-readable type code, the failing value, and the constraint that was violated. Use .annotate() to visualize errors in context.
js
const { error } = schema.validate(data, { abortEarly: false });
// error.details → [
// {
// message: '"age" must be greater than or equal to 18',
// path: ['age'],
// type: 'number.min',
// context: { limit: 18, value: 12, label: 'age', key: 'age' }
// }
// ]Type codes like string.email, number.min, any.required make programmatic error handling straightforward.