# Base32 Coding

## Overview

[`Base32`](/api/Base32) implements the raw BIP-173 character set — the alphabet Bech32 addresses
build on — without a checksum or human-readable prefix. Reach for it when a protocol hands you
bare base32 data; for checksummed addresses use [`Bech32m`](/api/Bech32m) instead.

## Recipes

### Encode & Decode Data

Round-trip hex payloads with [`Base32.fromHex`](/api/Base32/fromHex) and
[`Base32.toHex`](/api/Base32/toHex); [`Base32.fromBytes`](/api/Base32/fromBytes) and
[`Base32.toBytes`](/api/Base32/toBytes) cover `Uint8Array` payloads.

```ts twoslash
import { Base32 } from 'ox'

const encoded = Base32.fromHex('0xdeadbeef')
// @log: 'm6kmamc'
const decoded = Base32.toHex(encoded)
// @log: '0xdeadbeef'
```

## Best Practices

### Prefer Bech32m for Addresses

Base32 carries no integrity protection. When encoding anything a user might copy by hand — an
address, an invoice — use [`Bech32m`](/guides/data/bech32m), which adds a prefix and a
checksum that catches transcription errors.

## See More

<Cards>
  <Card icon="lucide:at-sign" title="Bech32m Coding" description="Checksummed, prefixed addresses built on this alphabet." to="/guides/data/bech32m" />

  <Card icon="lucide:hash" title="Work with Bytes & Hex" description="The primitive types every codec converts to and from." to="/guides/data/bytes-hex" />
</Cards>