CRC32 vs SHA-256 — which should I use?

Use CRC32 for accidental corruption detection in non-adversarial contexts (file transfer, storage). Use SHA-256 (or stronger) when an adversary could craft malicious data — CRC32 is trivial to forge. SHA-256 also has a much lower collision probability.

Can a checksum guarantee data is correct?

No. A matching checksum only means no detected change. Different inputs can share a checksum (collision). CRC32 collisions are common on large data; SHA-256 collisions are computationally infeasible but not impossible. For full assurance, use digital signatures.

Vérifier fichier ou données integrity avec un checksum

Calculer a CRC32 or SHA-256 checksum of data and compare it against a known value to detect corruption or tampering.

1

A checksum is a fixed-size digest of arbitrary data. If the data changes by even one bit, the checksum changes completely. Comparing checksums detects accidental corruption (CRC32) or intentional tampering (SHA-256).
2

Somme de contrôle CRC32 →
For fast, non-cryptographic integrity checks (file transfer, ZIP entries), compute CRC32. It is a 32-bit checksum — fast but can miss some corruptions and provides no security.
3

Calculatrice de hash →
For cryptographic integrity (detecting deliberate tampering), compute SHA-256. A 256-bit hash makes it infeasible to find a different input with the same hash.
4

Comparateur de hash →
Enter the computed hash and the expected hash. The tool compares them in constant time and reports match/mismatch — useful for automating verification scripts.

Questions fréquentes

CRC32 vs SHA-256 — which should I use?: Use CRC32 for accidental corruption detection in non-adversarial contexts (file transfer, storage). Use SHA-256 (or stronger) when an adversary could craft malicious data — CRC32 is trivial to forge. SHA-256 also has a much lower collision probability.
Can a checksum guarantee data is correct?: No. A matching checksum only means no detected change. Different inputs can share a checksum (collision). CRC32 collisions are common on large data; SHA-256 collisions are computationally infeasible but not impossible. For full assurance, use digital signatures.