What is an MD5 and SHA-256 hash: complete guide to cryptographic hash functions

A hash function is a mathematical algorithm that takes any amount of data as input and produces a fixed-length string as output. That string is called a hash, digest, or checksum.

The fundamental properties of a good hash function are:

• Deterministic: The same input always produces the same output. Hash "hello" a thousand times and you get the same hash a thousand times.
• One-way: It's practically impossible to obtain the original data from the hash. You can't "unhash."
• Avalanche effect: A minimal change in input produces a completely different hash. "hello" and "Hello" generate hashes that look nothing alike.
• Collision resistant: It's extremely difficult to find two different inputs that produce the same hash.

Concrete example:

Input	SHA-256 Hash
hello	2cf24dba5fb0a30e26e83b2ac5b9e29e1b161e5c1fa7425e73043362938b9824
Hello	185f8db32271fe25f561a6fc938b2e264306ec304eda518007d1764826381969
hello (again)	2cf24dba5fb0a30e26e83b2ac5b9e29e1b161e5c1fa7425e73043362938b9824

Notice how "hello" and "Hello" (only the capital changes) produce totally different hashes, but "hello" repeated gives the same result. Generate your own hashes with the NexTools hash generator.

MD5 (Message Digest Algorithm 5) was created by Ronald Rivest in 1991 and for years was the standard for file integrity verification. It produces a 128-bit hash (32 hexadecimal characters).

Why MD5 is no longer secure:

In 2004, researchers demonstrated that generating MD5 collisions was possible in seconds. This means two completely different files can have the same MD5 hash. In 2008, a team used this vulnerability to create a fake SSL certificate. In 2012, the Flame malware used MD5 collisions to masquerade as a Windows update.

Still-valid uses of MD5 in 2026:

• Non-cryptographic checksums: Verifying a download wasn't corrupted during transfer.
• Deduplication: Detecting duplicate files in a filesystem. Accidental collision probability is negligible.
• Cache keys: Generating fast keys for hash tables in non-cryptographic applications.

NEVER use MD5 for: Storing passwords, verifying digital signatures, or any security-critical context.

SHA-256 (Secure Hash Algorithm 256-bit) is part of the SHA-2 family, designed by the NSA and published by NIST in 2001. It produces a 256-bit hash (64 hexadecimal characters).

Why SHA-256 is the standard in 2026:

• No known collisions: After 25 years, no one has found a collision in SHA-256. The search space is 2^256 possibilities — more than the atoms in the observable universe.
• Bitcoin and blockchain: SHA-256 is Bitcoin's core hash function. Bitcoin mining consists of finding a nonce that produces a SHA-256 hash with a certain number of leading zeros. In 2026, the Bitcoin network computes approximately 600 quintillion SHA-256 hashes per second.
• SSL/TLS certificates: All modern HTTPS certificates use SHA-256 for signing. Since 2017, browsers reject certificates signed with SHA-1.
• Git: Git uses SHA-1 (migrating to SHA-256) to identify commits and objects.

Speed: SHA-256 is slower than MD5 (approximately 3-4x in software), which is actually an advantage for password storage — it makes brute force attacks more expensive.

Algorithm	Output bits	Hex length	Secure in 2026	Relative speed
MD5	128	32 chars	No (collisions in seconds)	Fastest
SHA-1	160	40 chars	No (collision demonstrated 2017)	Fast
SHA-256	256	64 chars	Yes	Medium
SHA-512	512	128 chars	Yes	Medium (faster on 64-bit)
SHA-3	256/512	64/128 chars	Yes	Slower

When to use each:

• MD5: Only for non-cryptographic checksums and deduplication
• SHA-1: Avoid. Only for legacy system compatibility
• SHA-256: Default choice for all cryptographic use
• SHA-512: When you need a longer hash or work on 64-bit architectures
• SHA-3: Alternative to SHA-2 with a different design. Use if regulations require it

Compare different algorithms with the NexTools hash generator supporting MD5, SHA-1, SHA-256, and SHA-512.

Hashes are in more places than you might think:

1. Password storage. Databases should NEVER store passwords in plain text. Instead, they store the hash. When you log in, the system hashes your password and compares with the stored hash. For passwords, specialized functions like bcrypt, scrypt, or Argon2 are used (adding "salt" and being intentionally slow).

2. File integrity verification. When downloading software (Linux ISOs, for example), the site publishes the file's SHA-256 hash. After downloading, you calculate the hash locally and compare. If they match, the file wasn't altered.

3. Digital signatures. Digitally signed documents use hashes. The document's hash is calculated, encrypted with the signer's private key, and attached. The recipient verifies by decrypting with the public key.

4. Blockchain and cryptocurrencies. Each block in a blockchain contains the previous block's hash, creating an immutable chain.

5. Version control (Git). Git identifies every commit, file, and directory by its SHA-1 hash.

6. Duplicate detection. Services like Google Drive or Dropbox detect duplicate files by comparing hashes instead of full content, saving space and bandwidth.

Several ways to calculate hashes depending on your needs:

Option 1: Online tool. The NexTools hash generator calculates MD5, SHA-1, SHA-256, and SHA-512 directly in your browser. No data leaves your computer.

Option 2: Terminal.

• Linux/Mac: echo -n "text" | sha256sum
• MD5: echo -n "text" | md5sum
• File: sha256sum file.zip
• Windows PowerShell: Get-FileHash file.zip -Algorithm SHA256

Option 3: JavaScript (Node.js).

const crypto = require('crypto');
const hash = crypto.createHash('sha256').update('text').digest('hex');

Option 4: Python.

import hashlib
hashlib.sha256(b'text').hexdigest()

Remember: a hash is NOT encryption. If you need to encrypt data, use algorithms like AES. Hashes are irreversible by design. Check our guide on Base64 and encoding to understand the difference between encoding, hashing, and encryption.

1. Rainbow tables. Pre-computed dictionaries of hash→text for millions of common passwords. If you store a plain SHA-256 hash of "password123," an attacker finds it in the table in milliseconds. Solution: Add a "salt" (unique random value per user) before hashing.

2. Brute force. Trying all possible combinations. With MD5, a modern GPU can try 10+ billion combinations per second. With SHA-256 it's ~3 billion, and with bcrypt only ~30,000. Solution: Use slow functions (bcrypt, Argon2) for passwords, and long passwords (12+ characters).

3. Length extension attacks. Affect MD5 and SHA-256 when used directly as MAC (Message Authentication Code). Solution: Use HMAC instead of direct hash.

4. Collisions. Finding two inputs with the same hash. Already trivial for MD5, demonstrated for SHA-1 (cost: ~$100,000 in 2017). Does not exist for SHA-256. Solution: Use SHA-256 or higher.

SHA-3 (Keccak). Approved by NIST in 2015, SHA-3 uses a completely different design from SHA-2 (sponge construction vs. Merkle-Damgard). It's not a replacement for SHA-2 (which remains secure) but an alternative in case a vulnerability is ever found in SHA-2.

Quantum threat. Quantum computers could use Grover's algorithm to halve hash security. SHA-256 would go from 256-bit to 128-bit equivalent security, which is still sufficient. SHA-512 would be equivalent to 256 bits, still very secure. Migrating to SHA-256+ is prudent even without an immediate quantum threat.

BLAKE3. A more modern, extremely fast hash (up to 10x faster than SHA-256 in software) that maintains high security. Not yet a NIST standard but gaining adoption in 2026, especially in data storage and deduplication.

For most uses in 2026, SHA-256 remains the right choice. Try different algorithms with the NexTools hash generator.