Really fast. Most of our search queries take from 1 to 20 milliseconds to process. That’s up to 200 times faster than our competitors.
That said, query performance depends on many factors, from the size of the data you’re searching through to how CPU-intensive a query is. For example, features like geo-sorting and distinct are more “expensive”.
Our ability to achieve unparalleled speed relies on a multitude of features, many of which we outlined in this blog post. For instance, the Algolia engine handles the vast majority of computation at indexing time, as opposed to query time. Beyond this, we control the full stack end-to-end. We’ve obsessed over every detail, from obtaining high-quality infrastructure (bare-metal servers) to crafting our own Ubuntu-based operating system, painstakingly modified for search engine performance.
To ensure queries are fast all around the world, we’ve also created our Distributed Search Network (DSN). This lets you automatically duplicate your search engine and data in multiple locations across the world. This is a complete copy of the search engine with all data, not just a cache. When your users perform a search, their queries reach the closest data center, significantly reducing the latency of the request.