JSON Generation
Generate structured JSON content based on prompts, messages, or custom schemas. Useful for creating API responses, structured records, or documents.
Basic JSON Generation
When you need to generate structured data with specific fields, use schema-based generation. This ensures the output follows your required format with proper data types. Learn more about basic JSON generation.
const {
data: { payload }
} = await unbody.generate
.json("Generate a product catalog entry for an e-commerce site:", {
schema: {
type: "object",
properties: {
productName: { type: "string" },
price: { type: "number" },
category: { type: "string" },
inStock: { type: "boolean" },
description: { type: "string" }
}
}
});Generation with Zod Schema
When you need type-safe JSON generation, use Zod schemas to define your output structure. This approach provides full TypeScript integration and runtime validation.
import { z } from "zod";
const {
data: { payload }
} = await unbody.generate
.json("Generate a product catalog entry for an e-commerce site:", {
schema: z.object({
productName: z.string()
.min(3, "Product name must be at least 3 characters")
.describe("Name of the product"),
price: z.number()
.positive("Price must be positive")
.describe("Current price in USD"),
originalPrice: z.number()
.positive()
.optional()
.describe("Original price before discount"),
category: z.enum(["Electronics", "Home & Kitchen", "Fashion", "Books"])
.describe("Product category"),
inStock: z.boolean()
.default(true)
.describe("Whether the product is in stock"),
description: z.string()
.min(50, "Description must be detailed")
.max(1000)
.describe("Detailed product description"),
tags: z.array(z.string())
.optional()
.describe("Product tags for search")
})
});Multi-Message Generation
When you need to generate structured data through conversation, use Message-Based Input with Schema. This helps when you need to provide context or specific instructions before generating the final structured output.
import { z } from "zod";
const {
data: { payload },
} = await unbody.generate
.json(
[
{
role: "system",
content: "You are a technical documentation specialist. Create API endpoint specifications."
},
{
role: "user",
content: "Generate documentation for a user registration endpoint."
},
],
{
schema: z.object({
endpoint: z.string()
.startsWith("/")
.describe("API endpoint path"),
method: z.enum(["GET", "POST", "PUT", "DELETE", "PATCH"])
.describe("HTTP method"),
parameters: z.array(
z.object({
name: z.string(),
type: z.enum(["string", "number", "boolean", "object", "array"]),
required: z.boolean().default(true),
description: z.string().optional(),
validation: z.record(z.string()).optional()
})
).min(1, "At least one parameter is required"),
responseCode: z.number()
.int()
.min(200)
.max(599)
.describe("HTTP response code"),
authentication: z.boolean()
.default(false)
.describe("Whether endpoint requires authentication"),
rateLimit: z.object({
requests: z.number().positive(),
period: z.string()
}).optional()
.describe("Rate limiting configuration")
}),
}
);Learn more about advanced JSON generation features in our JSON Generation API Guide.