Sphinx: Open source full text search server, designed with performance, relevance, and integration simplicity

Sphinx is an open source full text search server licensed under licensed under GPLv2. But it can also be commercially licensed, which is what you’ll need if you want to embed it in non-GPL products.

Written in C++, it works on Windows, Linux, MacOS, and several other systems like FreeBSD and Solaris. Its clusters can handle tens of billions of documents, along with hundreds of millions of search queries each day. It’s what powers websites like Craigslist, Groupon, MetaCafe, and Living Social.


  • Indexing capabilities: It can index up to 10 to 15 MB of text per second and that’s on a single CPU core. On a dedicated indexing machine, you’re looking at least 60 MB of text per second on every server.
  • Robust search: Let’s say you have a million documents that total 1.2 GB of data. With a 2-core desktop machine with 2 GB of RAM, you can power more than 500 queries per second.
  • Scalability: It can handle your data as your enterprise grows. The largest Sphinx cluster index contains more than 25 billion documents. It can also handle an impressive number of search queries. With Craigslist, it powers more than 300 million search queries each day.
  • 3 different APIs: With SphinxAPI, you have a native library available for languages like PHP, Java, Perl, C, Python, and others. SphinxSE is a storage engine you can plug in for MySQL. With SphinxQL, you can use standard MySQL client library and query syntax.
  • Improved relevance rankings: Better matches are ranked at the top, and you can even tweak the weights of the relevance functions.

Related Posts