Markdown Audit
Test Accept Markdown on any URL
Paste a public URL. We send Accept: text/markdown and Accept: text/html, then compare sizes and verify negotiation headers. Works on any site.
Advanced options
Credentials are sent only for the outbound check and are not stored.
Checks
What the audit verifies
Each URL is fetched twice — as HTML and as Markdown — then scored against the same contract accept-md implements.
- Markdown response
- HTTP 200 with Content-Type: text/markdown and a body that is Markdown, not HTML.
- HTML fallback
- Browser requests still get text/html — your pages stay unchanged.
- Vary: Accept
- Caches store separate representations so Markdown and HTML do not collide.
- Distinct representations
- Live Markdown differs from HTML, or the estimate shows meaningful size savings.
Results
How to read your report
Supported — the site returned real Markdown for Accept: text/markdown. All four checks should pass for production.
Not detected — no live Markdown response. The audit still shows an estimated conversion from HTML so you can see potential savings.
Expand Show response to inspect bodies side by side. For context, read what Accept Markdown means or troubleshooting.
FAQ
Common questions
How do I audit my website for Accept Markdown?+
Accept: text/markdown and Accept: text/html, then scores four production-readiness checks.What is Accept Markdown?+
Accept: text/markdown. Learn the protocol.Does this only work on accept-md sites?+
What does "estimated" mean in the results?+
Are my URLs or credentials stored?+
Can I audit multiple pages at once?+
Why might audits fail even when Markdown works?+
Vary: Accept, wrong Content-Type, or a CDN serving cached HTML for Markdown requests. Troubleshooting.Why use this instead of curl?+
What should I do if my site is not supported?+
npx accept-md init in your project, then re-run the audit. See the installation guide.CLI
Test from the command line
These three commands mirror what the audit does under the hood.
curl -s -H "Accept: text/markdown" https://example.com/pagecurl -sI -H "Accept: text/markdown" https://example.com/pagecurl -sI -H "Accept: text/html" https://example.com/page