---
title: "Your JSON Schema can be valid and still fail an LLM provider"
url: https://daily.dev/posts/your-json-schema-can-be-valid-and-still-fail-an-llm-provider-8klv9rybo
source_url: https://daily.dev/posts/your-json-schema-can-be-valid-and-still-fail-an-llm-provider-8klv9rybo
type: freeform
source: "Build With GenAI"
author: "Dan Mercede"
published: 2026-06-28T19:56:44.064Z
updated: 2026-06-28T19:57:01.337Z
reading_time: 2
upvotes: 2
comments: 0
language: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Your JSON Schema can be valid and still fail an LLM provider

**[Build With GenAI](https://daily.dev/sources/buildwithgenai)** · [@danmercede](https://daily.dev/danmercede) · 2 min read · 2 upvotes · 0 comments

## Summary

JSON Schema validation for LLM structured outputs is more complex than standard schema validation — a schema can be valid JSON Schema yet still be rejected by OpenAI, Anthropic, or Gemini due to provider-specific constraints like unsupported keywords, nesting limits, or enum edge cases. The post advocates treating structured-output schemas as production contracts and validating them against provider-specific rules before runtime, ideally in CI. SchemaFit is introduced as a tool designed to act as a schema compatibility gate for structured-output workflows, catching portability issues early across providers and tools like Instructor, BAML, Pydantic, and Zod.

## Content

Structured-output workflows have a weird failure mode:

A schema can be valid JSON Schema, pass local checks, look reasonable in code review… and still fail once it hits a provider API.

The problem is that “valid schema” and “accepted by OpenAI / Anthropic / Gemini structured-output constraints” are not always the same thing.

Examples I kept running into:

- unsupported schema keywords
- provider-specific nesting limits
- optional vs required field mismatches
- enum/object-shape edge cases
- schemas that pass local validation but fail at runtime
- schema drift between prompt code, app code, and CI

That pushed me toward a simpler pattern:

**Treat structured-output schemas like production contracts.**

Not just prompt helpers.

Not just type hints.

Not just Zod/Pydantic objects sitting near the API call.

Actual contracts that should be checked before runtime.

The workflow I wanted:

1. define the schema
2. lint it against provider-specific constraints
3. catch portability issues early
4. fail in CI before production API calls fail
5. keep the schema usable across providers where possible![SchemaFit Ensure compatible structured outputs.png](https://media.daily.dev/image/upload/s--StxZEO85--/f_auto/v1782676477/ugc/content_d92a9b25-c158-4ee6-b26d-02a2259d4e2f?_a=BAMAMicg0)

That is what I’m testing with **SchemaFit**:

[https://schemafit.danmercede.com/](https://schemafit.danmercede.com/)

It is meant to sit before structured-output API calls and act more like a schema compatibility gate than a full framework.

Where I think this fits:

- OpenAI / Anthropic / Gemini structured-output workflows
- CI checks for agent or LLM app schemas
- teams using Instructor, BAML, LiteLLM, Vercel AI SDK, Pydantic, or Zod
- systems where schema drift causes production issues
- provider-portability testing

Where it probably does **not** fit:

- quick prototypes
- one-off prompts
- single-provider projects with simple schemas
- teams already fully abstracted behind a higher-level framework

The broader point:

**Structured output is only reliable if the contract is validated before the model call.**

Curious how others are handling this.

Are you checking structured-output schemas before runtime, or mostly discovering provider-specific failures when the API rejects them?

## Similar posts on daily.dev

- [Structured Outputs in 2026: A Developer Guide](https://daily.dev/posts/structured-outputs-in-2026-a-developer-guide-4qqi5lmpc) · Alex CloudStar · 1 upvotes · 0 comments
- [Why JSON Schema matters more than ever in the age of generative AI](https://daily.dev/posts/why-json-schema-matters-more-than-ever-in-the-age-of-generative-ai-xskn2gm03) · The New Stack · 1 upvotes · 0 comments
- [Structured Outputs with LLMs: JSON Mode, Function Calling, and When to Use Each](https://daily.dev/posts/structured-outputs-with-llms-json-mode-function-calling-and-when-to-use-each-fmrab8cqw) · Towards Data Science · 4 upvotes · 2 comments

---

[View this post on daily.dev](https://daily.dev/posts/your-json-schema-can-be-valid-and-still-fail-an-llm-provider-8klv9rybo)