Skip to main content

What is Search?

Encrypting data has always come with one costly tradeoff: it makes searching over your data very complicated. Basis Theory’s search feature aims to make that as simple and painless as possible, without any need to detokenize or decrypt your data.

How It Works

Basis Theory utilizes the concept of a Blind Index, which is similar to an irreversible cryptographic hash, to empower fast searches across encrypted sensitive data. When a token is created, its data can be securely indexed to support exact-match searching on one or more patterns derived from the original plaintext data.

Token search indexes will be computed from the array of search index expressions specified within the request when creating a token. Some token types have default search indexes, but they can be overridden with custom indexes per-token to fit your use case.


In order to search Tokens using the API, you must use an Application with the token:search permission. The search results are filtered based on the permissions associated with the Application; you cannot search for tokens you do not have permission to access.

Query Syntax

Basis Theory uses a Lucene-based query syntax to power token search. See our docs on the Token Search Query Syntax for more details.


What is the difference between Search and Fingerprints?

Fingerprints are a measure of uniqueness, not a representation of the underlying data. Fingerprints can be used to locate duplicate data, for instance, but do not allow you to find specific data. You cannot find a card number ending in 4242 with fingerprints.

How does searching Tokens affect my Monthly Active Token usage?

Each Token that matches the search query made via the API and is returned in the result set becomes an Active Token for that month. Searching through the Portal as a logged in user will not affect MAT usage.