Open Graph tags control how your URLs appear when shared on social platforms. Getting them right is non-trivial: each platform has its own crawler, its own caching behavior, and its own rules for which tags it reads and ignores. Validating OG tags means simulating what each crawler sees — not what your browser renders — and checking that against each platform's documented constraints.

Why Browser DevTools Are Not Enough

When you inspect your page's <head> in Chrome DevTools, you are seeing the DOM after JavaScript has run. Social crawlers don't run JavaScript. They make a plain HTTP request, read the raw HTML, and extract meta tags from that. If your OG tags are injected by a React component, set by react-helmet, or updated dynamically after page load, crawlers may see nothing — even though DevTools shows the correct tags.

The correct way to validate OG tags is to fetch your page server-side, exactly as each crawler does, and inspect the HTML response before any JavaScript executes.

How Each Platform's Crawler Works

Twitter / X (Twitterbot)

—User-Agent: Twitterbot/1.0
—Reads twitter:* tags first, falls back to og:* tags.
—twitter:card is required — without it, Twitter may not render a card at all.
—Caches previews per URL. No official cache-purge tool; append a query string (?v=2) to force a fresh fetch.
—Does not execute JavaScript.

LinkedIn (LinkedInBot)

—User-Agent: LinkedInBot/1.0 (compatible; LinkedInBot/1.0; +http://www.linkedin.com)
—Reads only og:* tags — ignores all twitter:* tags.
—Aggressively caches for up to 7 days. Use the LinkedIn Post Inspector to force a re-crawl.
—Images below 200×200px are hidden entirely.
—Does not execute JavaScript.

Facebook (facebookexternalhit)

—User-Agent: facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)
—Reads og:* tags. Also reads article:* namespace tags for article og:type.
—Caches for ~30 days. Use the Facebook Sharing Debugger to force a re-crawl.
—Images below 200×200px are not shown; below 600×315px are shown as small thumbnails.
—Does not execute JavaScript.

Discord (Discordbot)

—User-Agent: Discordbot/2.0 (+https://discordapp.com)
—Reads og:* tags. Also reads theme-color meta tag for embed accent color.
—Supports og:video for inline video embeds.
—Re-fetches automatically after a few minutes. Paste the URL again to refresh.
—Does not execute JavaScript.

Slack (Slackbot)

—User-Agent: Slackbot-LinkExpanding 1.0 (+https://api.slack.com/robots)
—Reads og:* and twitter:* tags. Uses twitter:card to determine image size.
—Without twitter:card="summary_large_image", shows a small thumbnail rather than a large image.
—Workspace admins can disable link unfurling per-domain.
—Does not execute JavaScript.

What to Validate for Each OG Tag

og:title

—Present and non-empty
—Twitter: max 70 characters before truncation
—LinkedIn: max 150 characters
—Facebook: max ~100 characters in feed
—Avoid HTML entities — write "Don't" not "Don't"

og:description

—Present and non-empty
—Twitter: max ~200 characters
—LinkedIn: max ~300 characters
—Facebook: max ~300 characters in most surfaces
—Target 150 characters — safe across all platforms

og:image

—Absolute URL (starts with https://)
—Returns HTTP 200 with an image Content-Type
—Publicly accessible without authentication
—Minimum 200×200px for LinkedIn and Facebook; 300×157px for Twitter
—Recommended: 1200×630px (works on all platforms)
—Include og:image:width and og:image:height to prevent layout shift

og:url

—Should match the canonical URL of the page
—Platforms use this for deduplication — mismatched og:url can cause ghost entries or broken sharing counts

twitter:card

—Required for Twitter to render a card — without it, Twitter may show no preview
—"summary_large_image" for landscape images (1200×628), "summary" for square (400×400)
—Slack also respects twitter:card for image display mode

Platform-Specific Validation Tools

Each major platform provides an official debugger. These are useful for forcing cache re-crawls after you fix OG tags:

—Twitter/X — twitter.com/cards/validator (may require login). Alternatively, append ?v=2 to your URL and post again.
—Facebook — developers.facebook.com/tools/debug — shows what Facebook's crawler sees and lets you "Scrape Again" to force a re-fetch.
—LinkedIn — linkedin.com/post-inspector — enter any URL to see what LinkedIn will display and force a cache refresh.
—Discord — no official tool. Just paste the URL in a test server and wait a few minutes for Discord to re-fetch.
—Slack — no official tool. Paste the URL in a direct message to yourself.

TipPlatform tools only show you one platform at a time and require logging in to their respective services. For a single-page view of all six platforms at once — with validation errors and fix suggestions — use OGProof.

Manual Validation with curl

You can simulate any crawler with curl by setting the User-Agent header. This shows you exactly what the crawler receives, with no JavaScript execution:

# Fetch as Twitterbot and extract OG tags
curl -s -A "Twitterbot/1.0" "https://example.com/your-page" \
  | grep -i '<meta.*og:\|<meta.*twitter:'

# Fetch as LinkedInBot
curl -s -A "LinkedInBot/1.0" "https://example.com/your-page" \
  | grep -i '<meta.*og:'

# Fetch as facebookexternalhit
curl -s -A "facebookexternalhit/1.1" "https://example.com/your-page" \
  | grep -i '<meta.*og:'

# Fetch as Discordbot
curl -s -A "Discordbot/2.0" "https://example.com/your-page" \
  | grep -i '<meta.*og:\|<meta.*theme-color'

If the grep returns nothing for og:image, your tags are either missing from the server response or being injected by JavaScript after the initial HTML. If the URL returns a redirect, follow it with curl -L.

Common Validation Failures and Their Fixes

1.og:image missing from server HTML — tags are being set by JavaScript. Fix: use SSR (Next.js Metadata API, generateMetadata()).
2.og:image is an HTTP URL — upgrade to HTTPS. Twitter and iMessage will not display HTTP images.
3.og:image returns 403 to crawlers — the image host is blocking non-browser user agents. Fix: allow all user agents, or move the image to a public CDN.
4.twitter:card not set — Twitter renders a text-only card or no card at all. Fix: add <meta name="twitter:card" content="summary_large_image" />.
5.og:url points to a different domain — some platforms use og:url instead of the actual URL for deduplication. Fix: set og:url to the exact canonical URL.
6.og:description over 150 characters — will be truncated differently per platform. Fix: write a 100–150 character description.
7.Image ratio wrong for Twitter — you have a square image but twitter:card="summary_large_image". Fix: switch to twitter:card="summary" or use a 2:1 landscape image.

Automating OG Validation in CI

For teams that deploy frequently, running a manual validation check after every deploy is impractical. You can automate basic checks with a simple script that runs in your CI pipeline:

#!/bin/bash
# validate-og.sh — run after deploy to check critical OG tags
URL="$1"

if [ -z "$URL" ]; then
  echo "Usage: ./validate-og.sh https://example.com/page"
  exit 1
fi

HTML=$(curl -s -A "Twitterbot/1.0" "$URL")

check_tag() {
  local tag="$1"
  local label="$2"
  if echo "$HTML" | grep -qi "$tag"; then
    echo "✓ $label"
  else
    echo "✗ $label MISSING" && FAILED=1
  fi
}

check_tag 'og:title' 'og:title'
check_tag 'og:description' 'og:description'
check_tag 'og:image' 'og:image'
check_tag 'og:url' 'og:url'
check_tag 'twitter:card' 'twitter:card'

if [ "$FAILED" = "1" ]; then
  echo "OG validation failed."
  exit 1
fi
echo "All OG tags present."

This script only checks for presence — it does not validate image dimensions, HTTPS, or truncation. For full validation including per-platform rendering, the OGProof API provides a programmatic endpoint that returns structured validation results for all six platforms.

OGProof: Validate All Six Platforms at Once

OGProof was built specifically to solve the multi-platform validation problem. Paste any URL and it fetches your page server-side (with no JavaScript execution), parses all OG, Twitter Card, and meta tags, and renders accurate previews for all six platforms: Twitter, LinkedIn, Facebook, Discord, Slack, and iMessage.

—Server-side fetch — sees exactly what each crawler sees, not what your browser renders
—Per-platform validation — each tag is checked against each platform's documented constraints
—Visual previews — see the rendered card, not just raw tag values
—Fix suggestions — specific, actionable errors when tags are missing or invalid
—Free to start — validate any public URL, no account required

Open Graph Validator: How to Test OG Tags Across All Platforms