The latest articles on DEV Community by sravan27 (@sravan27). https://dev.to/sravan27 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3887565%2Fa1b9fa16-7e94-48fd-b960-fc7c05b0596f.png https://dev.to/sravan27 en sravan27 Sun, 19 Apr 2026 16:22:26 +0000 https://dev.to/sravan27/your-structured-outputs-are-probably-less-portable-than-you-think-2p1l https://dev.to/sravan27/your-structured-outputs-are-probably-less-portable-than-you-think-2p1l <p>Structured outputs are less portable across LLM providers than they look.</p> <p>A schema change can seem harmless and still break one provider while passing another.</p> <p>I built Schema Gateway for that exact failure mode.</p> <p>What it does:</p> <ul> <li>compile one schema into provider-ready request shapes</li> <li>diff a baseline schema against a candidate schema</li> <li>lint for portability issues</li> <li>normalize payloads against a schema</li> </ul> <p>It’s built for people working across providers like OpenAI, Gemini, Anthropic, and Ollama, where “structured output support” sounds similar until the edge cases show up.</p> <p>There’s an open-source/local path, plus a hosted Starter Access if you want to try the hosted API quickly.</p> <p>Main thing I’m looking for:<br> feedback from people actually shipping structured outputs or comparing providers in production. I’m especially interested in whether this pain is strong enough to justify a dedicated guardrail in real workflows.</p> <p>Live here:<br> <a href="https://schema-gateway.sridharsravan.workers.dev/" rel="noopener noreferrer">https://schema-gateway.sridharsravan.workers.dev/</a></p> ai devtools programming webdev