Controlled technical evidence

Inspect the response difference, not a ranking promise.

These dated checks use Prerender Buddy-owned routes. They compare the thin HTTP shell sent to a normal command-line request with the complete HTML returned through the configured crawler path.

Tested 2026-07-13 on 3 public routes and 5 supplied crawler user agents.

Controlled user-agent requests verify configured routing. They do not authenticate Google, Microsoft, OpenAI, Anthropic, or Perplexity.

1,698 B

Normal command-line response: the shared Vite application shell.

52-96 KB

Crawler response range across the three tested public routes.

HTTP 200

Route-specific titles, H1 headings, canonicals, and JSON-LD were present.

Route evidence

Three routes, one repeatable difference.

Byte counts describe the complete HTTP response. Visible-word, heading, and link counts use a simple HTML extraction and include shared navigation and footer content.

Crawler support

/crawler-support

HTTP 200

Normal HTTP shell

1,698 B

Crawler response

95,679 B

Visible words
2,458
H1-H3 headings
46
Links
93
JSON-LD blocks
1

Title: Crawler Support | Prerender Buddy

H1: Exact request agents, with the status explained.

Canonical: https://prerenderbuddy.com/crawler-support

Prerender.io comparison

/compare/prerender-io

HTTP 200

Normal HTTP shell

1,698 B

Crawler response

52,060 B

Visible words
1,142
H1-H3 headings
17
Links
85
JSON-LD blocks
1

Title: Prerender Buddy vs Prerender.io | Practical Comparison

H1: Prerender Buddy vs Prerender.io

Canonical: https://prerenderbuddy.com/compare/prerender-io

Migration guide

/migrate-from-prerender-io

HTTP 200

Normal HTTP shell

1,698 B

Crawler response

54,908 B

Visible words
1,329
H1-H3 headings
32
Links
76
JSON-LD blocks
1

Title: Migrate from Prerender.io | Prerender Buddy

H1: Migrate from Prerender.io without guessing at the request path.

Canonical: https://prerenderbuddy.com/migrate-from-prerender-io

Crawler-style routing check

The same rendered document reached five test agents.

Target route: /crawler-support

Googlebot

HTTP 200

Supplied UA: Googlebot/2.1

95,679 B rendered response

Bingbot

HTTP 200

Supplied UA: Bingbot/2.0

95,679 B rendered response

GPTBot

HTTP 200

Supplied UA: GPTBot/1.0

95,679 B rendered response

ClaudeBot

HTTP 200

Supplied UA: ClaudeBot/1.0

95,679 B rendered response

PerplexityBot

HTTP 200

Supplied UA: PerplexityBot/1.0

95,679 B rendered response

What this demonstrates

  • Configured crawler-style requests received substantially more complete HTML than the normal command-line shell.
  • The rendered response included route-specific titles, headings, canonicals, links, and structured data.
  • Five supplied user-agent families reached the configured rendered-response path.

What this does not demonstrate

  • The requests do not authenticate the crawler operators or prove that their production crawlers visited.
  • The snapshot does not prove indexing, rankings, AI citations, recommendations, traffic, or conversion changes.
  • Results from these owned routes should not be treated as a performance benchmark for every customer website.

Method

How to reproduce the check.

  1. 1Request each public URL normally and record status, final URL, content type, and response bytes.
  2. 2Repeat with a supplied crawler-style user agent and follow the canonical www-to-root redirect.
  3. 3Inspect the crawler response for title, H1, canonical URL, JSON-LD, headings, links, and visible text.
  4. 4Record the date and keep operator authentication separate from controlled routing verification.

Test your own public page

Use the same evidence path on the website that matters.

Start with a raw-versus-rendered check, then review exact crawler families and setup options if the initial HTML exposes limited visible content.

Run the comparison See crawler support