# Before an agent touches your docs, check the benchmark.

This public documentation benchmark tests whether an agent can discover the right page, fetch readable text, follow the instructions, and recover when the docs get fuzzy.

Use it as a public docs leaderboard, a fast AI-readability audit, or a way to compare API docs, developer docs, and help centers before a prospect or agent does.

Run a live benchmark

## Create a permanent report for your docs.

Every completed run is saved as a detailed report and added to the public docs benchmark leaderboard.

### Try one of these public docs URLs:

- docs.stripe.com
- render.com/docs
- docs.sentry.io
- docs.docsalot.dev

The crawl runs here, then opens the saved report with scores, failed checks, and recommended fixes.

## Search benchmark reports and open the full breakdown.

Compare public docs across categories, inspect the strongest performers, and jump into the detailed report for any site.

### Categories tracked segments

- All: 279
- DocsAlot runs: 123
- Other: 57
- AI: 40
- Productivity: 10
- Infrastructure: 21
- DevTools: 16
- Payments: 12

### Showing 1 - 25 of 279 matching docs.

| Docs                           | Category | Checks     | Warnings | Failed | Score  |
|--------------------------------|----------|------------|----------|--------|--------|
| [developer.nylas.com](/content/benchmarks/docs/developer-nylas-com/index.html)  | DocsAlot runs | 17/17      | 0        | 0      | Excellent|
| [runzivo.dev](/content/benchmarks/docs/runzivo-dev/index.html)                | DocsAlot runs | 17/17      | 0        | 0      | Excellent|
| [docs.eco.com](/content/benchmarks/docs/docs-eco-com/index.html)              | DocsAlot runs | 16/17      | 0        | 1      | Excellent|
| [docs.cdp.coinbase.com](/content/benchmarks/docs/docs-cdp-coinbase-com/index.html)| DocsAlot runs | 16/17      | 0        | 1      | Excellent|
| [Snowflake](/content/benchmarks/docs/snowflake/index.html)                    | Other       | 20/22      | 1        | 0      | A+     |
| [Promptless](/content/benchmarks/docs/promptless/index.html)                  | AI         | 20/22      | 1        | 0      | A+     |
| [Alchemy](/content/benchmarks/docs/alchemy/index.html)                        | Other       | 20/23      | 2        | 0      | A+     |
| [Bolt](/content/benchmarks/docs/bolt/index.html)                              | Productivity | 18/22      | 2        | 1      | A      |
| [ID.me](/content/benchmarks/docs/id-me/index.html)                            | Other       | 20/23      | 1        | 1      | A      |
| [Introduction](/content/benchmarks/docs/introduction/index.html)              | Other       | 18/22      | 2        | 0      | A      |
| [Backboard](/content/benchmarks/docs/backboard/index.html)                    | AI         | 18/18      | 0        | 0      | A+     |
| [Feather](/content/benchmarks/docs/feather/index.html)                        | Other       | 18/22      | 2        | 1      | A      |
| [Runpod](/content/benchmarks/docs/runpod/index.html)                          | Infrastructure | 19/22      | 1        | 1      | A      |
| [Merge](/content/benchmarks/docs/merge/index.html)                            | Other       | 21/23      | 0        | 1      | A      |
| [Nvidia Dynamo](/content/benchmarks/docs/nvidia-dynamo/index.html)            | Infrastructure | 18/23      | 3        | 0      | A      |
| [Fern](/content/benchmarks/docs/fern/index.html)                              | DevTools   | 21/23      | 0        | 1      | A      |
| [maesn](/content/benchmarks/docs/maesn/index.html)                            | AI         | 16/22      | 1        | 4      | A      |
| [Brevo](/content/benchmarks/docs/brevo/index.html)                            | Other       | 20/23      | 1        | 1      | A      |
| [Nvidia](/content/benchmarks/docs/nvidia/index.html)                          | Infrastructure | 17/23      | 3        | 1      | A      |
| [Skyflow](/content/benchmarks/docs/skyflow/index.html)                        | Infrastructure | 16/22      | 2        | 2      | A      |

# Why it matters

## Docs quality now affects AI usability.

This is not a vanity score. It is a practical read on whether an agent can discover, read, and act on your documentation.

1. ### Agents already use docs as product instructions

Coding agents, AI search products, and automated support flows read docs long before a human asks for help. If the docs are hard to fetch or parse, the product feels harder to use.

2. ### Docs quality now affects activation

Agent-readable docs shape implementation speed, support load, and whether a product feels trustworthy during evaluation.

3. ### Benchmarks make docs quality comparable

A shared benchmark gives teams a concrete way to compare docs quality, spot weak areas, and track improvements over time.

## Methodology

### How the benchmark scores docs.

This scoring model is informed by [AFDocs](https://afdocs.dev/) and adapted into a public benchmark teams can inspect, compare, and rerun.

1. / Discovery **Content Discoverability** 6 checks Can an agent find the right starting point?
   - **LLMS TXT exists** An llms.txt file is published at a predictable docs path.
   - **LLMS TXT is structured** The llms.txt file follows a format agents can parse quickly.
   - **Docs entry point is obvious** The primary docs home is easy to identify.
   - **Docs sitemap exists** The crawl map is present and reachable.
   - **Key docs links resolve** The sampled entry links resolve without dead ends.
   - **Public indexing path is stable** Agents can stay on a stable public path while crawling the docs.

2. / Markdown **Markdown Availability** 2 checks Can an agent fetch a clean text version of the docs?

3. / Page size **Page Size and Truncation Risk** 4 checks Can an agent read the page without losing important context?

4. / Structure **Content Structure** 3 checks Can an agent understand how the page is organized?

5. / URL stability **URL Stability and Redirects** 2 checks Can an agent trust that URLs will stay predictable?

6. / Observability **Observability and Content Health** 3 checks Can an agent tell whether the docs are fresh and trustworthy?

7. / Access **Authentication and Access** 2 checks Can an agent access the docs without getting blocked?