Data, record size, and usage limits

Size limits

Record size limits

Records can’t go beyond a certain size limit. This limit might depend on your plan—see the Algolia pricing page for more details. If you try to index a record that exceeds the limit, Algolia returns the Record is too big error.

Break up your records into smaller ones if needed.

Index size limits

The maximum size for a single index is 100 GB (or 1 GB for the Build plan). The number of records an index can have is only limited by the capacity of the hardware.

Special infrastructure options let you go much further than this limit. If you would like to know more, contact your account manager or the Algolia support team.

Application size limits

For regular plans and infrastructure, the maximum size for an application (cumulative size of its indices) is 100 GB (or 1 GB for the Build plan). This represents 80% of the RAM capacity of regular servers (128 GB), which leaves 20% of the RAM capacity to handle your indexing tasks. If the application size exceeds the 100 GB (or 1 GB for the Build plan) capacity, performance degrades severely: data swaps back and forth between temporary and permanent memory, which is a costly operation.

You can monitor your application size in the API Monitoring section.

Special infrastructure options let you go much further than this limit. If you would like to know more, contact your account manager or the Algolia support team.

Indexing usage limits

Maximum indexing operations

Algolia counts the number of operations performed every month. When you hit your plan’s limit, you’re charged for the extra operations, based on your plan’s over-quota pricing.

Indexing rate limit

Algolia delays or rejects indexing operations whenever a server is overloaded. If Algolia determines that indexing operations can negatively impact search requests, it takes action to favor search over indexing. This rate limit exists to protect the server’s search capacity.

Query usage

Algolia counts a search operation whenever you perform a search. In search-as-you-type implementations, this happens on every keystroke. If you’re querying several indices at each keystroke, then one keystroke triggers as many operations as queried indices, unless you use the multipleQueries method to do this.