2¹²⁸ Seed security, visualized
A visual BIP39 explainer

12 vs 24 words

Why twelve correctly generated words can already match Bitcoin’s practical cryptographic security—and what twenty-four words do and do not add.

The short answer: 12 words are sufficient when they contain a genuine 128 bits of uniform entropy. Twenty-four words increase seed-search space, but do not make Bitcoin’s elliptic curve stronger.
128
bits of security
2²⁵⁶ keysBitcoin private-key space
2¹²⁸ workBest known generic curve attack
Start here

Three numbers people mix up

Raw key length, attack difficulty, and mnemonic entropy are related—but they are not the same measurement.

2²⁵⁶

Private-key space

Bitcoin uses secp256k1. A private key is essentially a number selected from a set close to 2²⁵⁶ possibilities.

2¹²⁸

Curve-attack work

Elliptic-curve attacks have a square-root shortcut. Searching a 256-bit curve takes roughly 2¹²⁸ operations, not 2²⁵⁶.

2¹²⁸

Valid 12-word seeds

Twelve BIP39 words encode 128 random bits plus a 4-bit checksum: 2¹²⁸ valid possibilities.

The useful comparison is security strength, not raw key length.
A 256-bit elliptic-curve key is commonly described as providing about 128 bits of classical security.
Interactive anatomy

Where the bits live

Every BIP39 word selects one of 2,048 entries. Since 2,048 = 2¹¹, each word represents 11 bits.

Illustrative public words

12 words × 11 bits

132 total bits
128 entropy
4 CS
Random entropyChecksum
Random entropy
128 bits
2¹²⁸ possible seeds
Checksum
4 bits
Error detection, not additional randomness
Raw word bits
132 bits
12 × 11 = 132
The square-root shortcut

256-bit keys, 128-bit work

The best known generic elliptic-curve attack does not try every private key. Its work grows approximately with the square root of the key space.

2²⁵⁶
possible private keys
√ →
Pollard-rho style square-root attack
2¹²⁸
rough operations required
Bitcoin does not have only 2¹²⁸ private keys.
It has close to 2²⁵⁶ private keys; the strongest known generic classical attack is why we call its security roughly 128-bit.
Think like an attacker

Attackers choose the cheaper door

Increasing one barrier beyond the other does not raise the security of the complete system. Switch between 12 and 24 words to see the bottleneck.

Door A

Guess the seed

2¹²⁸

A valid 12-word mnemonic carries 128 random bits.

Door B

Reverse a public key

≈ 2¹²⁸

Once an applicable public key is available, generic secp256k1 attacks remain around 2¹²⁸ work.

Effective classical ceiling
min(128, 128) = 128 bits
Both doors are similarly hard
Scale simulator

How long is 2¹²⁸?

Move the slider. Even absurd guessing speeds barely make a dent. These figures are simple division, not estimates of practical wallet-testing performance.

Hypothetical attacker
10¹² / second

For perspective, this deliberately lets you choose fantastical speeds up to 10³⁰ operations per second.

Exhaust 2¹²⁸ possibilities
1.08 × 10¹⁹ years
Average success would take about half this time.
Exhaust 2²⁵⁶ possibilities
3.67 × 10⁵⁷ years
Vastly larger, but not Bitcoin’s weakest classical door.
The observable universe is roughly 1.38 × 10¹⁰ years old. At 10¹² attempts per second, exhausting 2¹²⁸ takes around 780 million universe-lifetimes.
For physical SeedTabs

The last word is partly random

The phrase “checksum word” is convenient but incomplete. The final word contains both remaining entropy and checksum bits.

12-word mnemonic

Draw 11 words, then add 7 random bits

11 × 11 = 121; 128 − 121 = 7

Eleven uniform tab draws supply 121 bits. A trusted offline workflow must add 7 uniformly random bits, then compute the 4 checksum bits that complete word 12.

The final bitstream
121bits from tab draws
7more random bits
4checksum bits
DrawnStill randomCalculated checksum
Do not simply choose one deterministic “valid last word” and assume you retained full entropy. For 12 words, select uniformly among the 128 valid final-word candidates; for 24 words, select uniformly among the 8 candidates—or use a trusted offline device that explicitly supplies the missing random bits.
Practical conclusion

Choose the workflow you can execute correctly

The best theoretical number loses to a weak random source, an exposed backup, or a restoration mistake.

24
Optional margin

More seed-search space

  • 256 bits of genuine entropy
  • Raises direct mnemonic brute force from 2¹²⁸ to 2²⁵⁶
  • Does not strengthen secp256k1 beyond its roughly 128-bit classical level
  • More words to back up correctly and keep private

Neither word count fixes…

A photographed seedA weak or broken RNGA leaked digital copyA forgotten passphraseBad checksum handlingPhishing or coercion
Bottom line: Use 12 words when you can confidently produce all 128 random bits and finish BIP39 correctly offline. Use 24 when your wallet or policy requires it, or when you value extra mnemonic-search margin enough to accept the backup burden.
Check your intuition

Three quick questions

No seed phrases required—this demo never asks for secret material.

1. Approximately how many Bitcoin private keys are there?

2. How much random entropy is encoded by a valid 12-word BIP39 phrase?

3. After 11 uniform SeedTab draws, how many random bits are still needed for full 12-word strength?

Read the originals

Sources and boundaries

This explainer is educational. It is not a wallet, seed generator, or substitute for your signing device’s documented offline procedure.

BIP39 specification

Defines allowed entropy lengths, checksum construction, 11-bit word indices, and mnemonic lengths.

NIST SP 800-57 Part 1

General guidance relating elliptic-curve key sizes to security strength; 256-bit ECC corresponds broadly to 128-bit strength.

SeedTabs repository

Laser-ready SVGs for a physical set of all 2,048 English BIP39 words, plus verification tooling and safety guidance.

Never paste a real seed phrase into this or any webpage. This file is self-contained and performs only educational calculations, but secret recovery material belongs exclusively in a trusted offline workflow.