Home / SearchWebCode vs searchcode
Alternatives
SearchWebCode vs searchcode
Provider descriptions link to public sources · current index built September 2026
SearchWebCode data: 131.2M indexed homepages, one homepage per hostname, including subdomains. The latest quarterly source snapshot is September 2026. Query dates describe checks against that stored index, not live installation checks. Data and methods.
People reach searchcode and SearchWebCode with the same instinct, "find this snippet everywhere", and the two tools look in different places. One reads source files committed to public repositories. The other reads the HTML, inline JavaScript and CSS that websites actually serve to visitors. The snippet you are holding usually belongs to one world or the other, and that decides the tool.
searchcode vs SearchWebCode: repository code vs live website HTML
searchcode.com was built by Ben Boyter as a free index of public source code, and its site still describes itself as "built on a decade of indexing billions of lines of code". Checked on October 28, 2026, the homepage leads with "code intelligence for LLMs": paste a GitHub URL for a dossier of languages, complexity, tech stack and code quality, or add its MCP server so an AI coding assistant can run code_search, code_analyze and file retrieval over any public git repository. It is free during its beta with no sign-up or API key, and a self-hosted "searchcode enterprise" edition indexes an organisation's own repositories on its own hardware. The classic results page and the old /api/codesearch_I/ endpoint answered 404 when we checked, and a query typed into the homepage returns the homepage. searchcode.com and its llms.txt describe the current product.
SearchWebCode searches 131.2M indexed homepages, one homepage per hostname, including subdomains, captured in a quarterly source snapshot. A query is a literal string, a boolean expression or a regular expression, every result is checked against the stored page, and each opens as the full source with the matches highlighted. There is no repository view, because websites are not repositories: what it finds is the markup and the scripts a site serves, including third-party tags, tracking IDs and embed codes that never sit in anyone's public repo.
At a glance
| Capability | SearchWebCode | searchcode |
|---|---|---|
| What is indexed | stored HTML, inline JavaScript and CSS of 131.2M homepages | files in public git repositories, read on demand per repository |
| Unit of a result | a hostname, with its full stored source | a file or symbol inside one repository |
| Query types | literal, boolean AND/OR/NOT, regex, TLD filter | full-text with boolean, regex and structural filters, scoped to a repository (vendor description, October 28, 2026) |
| Access | web app and REST API | web dossiers, MCP server and REST mirror of its tools |
| Pricing | pay-as-you-go at 5 cents per search credit, or Research/Pro plans | free during beta; enterprise edition licensed |
| Built for | technographics, OSINT pivots on identifiers, adoption counts, domain exports | understanding, auditing and searching a codebase, especially from an AI assistant |
When to use each
Reach for searchcode, or another repository search engine, when you want to see how a library or API is used in real projects, you need the file and the surrounding code rather than a page, you are evaluating or auditing a dependency, or you are working inside an AI coding assistant and want it to read a repository without cloning it.
Reach for SearchWebCode when the question is which websites deploy something: a tracking ID, a widget embed, a CDN path, a framework marker, a plugin folder. Its results are domains you can export, and its evidence is the page as served, so it covers code that was never published in a repository, from a marketing team's tag manager snippet to a template vendor's footer.
Other code search engines (grep.app, Sourcegraph, GitHub code search, PublicWWW)
- GitHub code search covers GitHub's public repositories and requires signing in. It supports exact strings in quotes, AND/OR/NOT, regular expressions between slashes, and qualifiers such as
repo:,path:,language:andsymbol:(GitHub's syntax documentation, checked October 28, 2026). - grep.app runs fast regex searches across a large set of public GitHub repositories with no account, filtered by repository, path and language. Good for "who calls this function" when you do not know which repository to open.
- Sourcegraph is code search and navigation for engineering teams over their own repositories, with a public search page for open-source code. Regex and structural search, symbol navigation and batch changes are its strengths; it is a product you adopt, not a lookup you run once.
- PublicWWW is the closest neighbour to SearchWebCode: a website source-code search with snippets and regex extraction. The PublicWWW comparison covers the differences in detail.
Example: find a snippet across websites
Take highlight.js. Its README tells you to call hljs.highlightAll() after loading the library. In a repository search that line turns up the library's own source, its documentation and thousands of projects that copied the README. On SearchWebCode, hljs.highlightAll returns the websites whose homepage HTML contains the call, about 9,000 of them in a probe on October 28, 2026 (an estimate), with developer documentation sites and package pages at the top. Open one and the call is highlighted in the stored source; export the hostnames and you have a deployment list no repository can give you. The same move works for any embed: plausible.io/js/script.js is Plausible's snippet, on 40,924 estimated matches, and a vendor's loader URL works the same way.
The reverse example matters just as much. To learn what highlightAll() does, or to find a bug in it, SearchWebCode is the wrong tool: it holds the minified call as it appears on a page, not the implementation. That is a repository question, and the repository tools above answer it.
How to search the source code of the web → Run the example →
Frequently asked questions
Is searchcode still a code search engine?
On October 28, 2026 searchcode.com described itself as code intelligence for LLMs: repository dossiers, an MCP server and a REST API that analyse and search any public git repository. Its classic search results page and the old codesearch API returned 404s. Check the site for the current state; this page records what was published on the date given.
Can searchcode search the HTML of a website?
No. searchcode works on code in public git repositories. The HTML, scripts and tags a website serves are not in a repository unless the site owner publishes them, so a question about live websites needs a website source index such as SearchWebCode.
How do I find a keyword in a website's source code?
For one site, open the page, press Ctrl+U (Cmd+Option+U on a Mac) and search the source with Ctrl+F. For every indexed site at once, type the keyword into SearchWebCode: it returns the homepages whose stored HTML, JavaScript or CSS contains it, each with the full source and the matches highlighted.
Is SearchWebCode free?
It has a free trial, then charges pay-as-you-go at 5 cents per search credit or through Research and Pro subscriptions with monthly allowances. Reading the stored source of a result costs nothing extra. Current terms are on the pricing page.
Which tool finds sites that share a tracking ID?
SearchWebCode. Tracking and analytics IDs live in page HTML, not in repositories, so search the literal ID and every indexed homepage that prints it comes back. The guide on finding websites sharing a Google Analytics ID walks through it.
Does either tool index private code?
No. searchcode reads public git repositories, and SearchWebCode reads the public homepages of websites. Private repositories and pages behind a login are outside both.