Base91 Encoder and Decoder

Pack bytes efficiently into 91 printable symbols

BasE91 uses a 91-character alphabet and variable groups of 13 or 14 bits, producing less overhead than Base64 for many byte sequences. This tool encodes UTF-8 bytes with the standard basE91 bit-packing rule and reverses the process with character validation. It is useful for interoperating with systems that explicitly expect basE91, not merely any printable-text encoding.

The default word test encodes to fPNKd, and decoding fPNKd returns the original four letters. Unlike fixed-block encodings, basE91 chooses 13 bits when the low value exceeds 88 and otherwise consumes 14, then emits two base-91 digits. The final partial buffer may emit one or two characters, which is why output length is not a fixed block multiple.

Use basE91 as encoding rather than encryption

BasE91 provides no secrecy, integrity, or authenticity; anyone can decode it, and a changed symbol can silently alter bytes. Decoded bytes must also form valid UTF-8 for this text-focused page. Every transformation runs in your browser without network calls. Preserve the exact alphabet and case when exchanging values, because similarly named Base91 variants can use different symbol ordering.