$0 · 100% browser-side

JSON Schema validator

Validate JSON syntax, check payloads against JSON Schema rules, and audit OpenAI / LLM Structured Outputs (strict: true) requirements locally in your browser.

Create API Key →

1. JSON Schema (or response_format)

2. JSON Payload to Validate

Schema Validation Result

Valid JSON & Schema

Payload syntax is valid JSON and satisfies all properties, types, enums, and required constraints in the schema.

Structured Outputs (strict: true) Audit

strict: true ready

Passes OpenAI Structured Outputs strict checks: root is object, every object sets additionalProperties: false, and every property is in required.

Run Structured Outputs on GPT, Claude, and Gemini

Use one KeelRouter API key on /v1/chat/completions or /v1/responses to test strict JSON schemas across multiple models without separate billing accounts.

Frequently asked questions

What does OpenAI Structured Outputs strict: true require?
Every object schema must have "additionalProperties": false, and every key inside "properties" must appear in "required" (use type: ["string", "null"] for optional fields).
Can I paste a full response_format wrapper or raw JSON Schema?
Yes. The validator accepts both raw JSON Schema objects (Draft 2020-12 / Draft-07 subset) and full OpenAI response_format: { type: "json_schema", json_schema: { ... } } payloads.
Does any JSON or schema data leave my browser?
No. Syntax parsing, JSON Schema validation, strict: true auditing, and prettifying run 100% client-side in your browser.
Which models on KeelRouter support Structured Outputs?
GPT-6.1 / GPT-5 / GPT-4o models, Gemini 3.x / 4.x, and Claude models via tool-use / JSON schema output all work through KeelRouter’s /v1/chat/completions and /v1/responses endpoints.