Skip to main content
ToolsFree.io🇪🇸es
By MD Tech Engineering SL··Updated ·4 min read

Strong Passwords: Length and Randomness

Share:𝕏LinkedIn

A password generator solves one narrow problem: producing a value that is less predictable than a phrase a person invents. Account security still depends on using a different password for every service, storing it safely, and protecting account recovery and sign-in with additional controls.

How the ToolsFree generator works

The password generator builds the selected character set and uses crypto.getRandomValues in your browser to choose characters. It does not upload the generated value to ToolsFree. The displayed entropy is a theoretical estimate that assumes independent, uniformly selected characters from that set; it is not a prediction of a real attack time.

Length, randomness, and uniqueness

More independently generated characters increase the search space. Reuse removes much of that benefit because one service breach can expose credentials that attackers try elsewhere. Prefer the longest generated password the destination accepts, make it unique to that account, and avoid editing it into a memorable pattern.

Search space, in bits of entropy8 chars, a–z≈38 bits12 chars, a–z A–Z 0–9≈71 bits16 chars, all ASCII≈105 bits
Length buys far more entropy than exotic characters do: each extra character multiplies the whole search space.

What we measured in its output

Claiming a generator is random is easy. Every generator claims it. So we counted: 2,000 passwords at the longest setting the tool offers, 256,000 character draws in all, taken from the generator's own bulk mode and tallied by a script.

Passwords generated2,000 x 128
Distinct among them2,000
Character draws counted256,000
Symbols in the alphabet88
Expected times each appears2,909
Observed range2,795 – 3,035
Spread a uniform source gives± 54
Widest deviation seen2.3 standard deviations
A sample taken from this generator's own bulk mode, counted by a script. Every figure below is observed, not modelled. Measured on 2026-08-29 with Chromium 151.0.7922.173. Re-running draws a fresh sample, so the counts move; what should not move is how far they spread.

The alphabet has 88 symbols, so over 256,000 draws each should turn up about 2,909 times. The rarest appeared 2,795 times and the commonest 3,035. That gap looks large until you work out what a flat source would do: the expected spread is about ±54, which puts those extremes 2.2 and 2.6 standard deviations from the middle. Across 88 symbols, an extreme of roughly that size is what you should expect to see. A result clustered much tighter than this would be the suspicious one.

One detail in the implementation is worth naming, because it is the part most generators get wrong. Turning a random 32-bit number into a symbol by taking the remainder is not uniform: 232 divided by 88 leaves 48 left over, so the first 48 symbols of the alphabet would come up very slightly more often than the other 40 — about one draw in 48.8 million. This generator discards the values in that incomplete final block and draws again, which removes the skew entirely rather than making it small. The measurement above is what that looks like from the outside.

What the sample does not show is any repeat: all 2,000 passwords were distinct. That is the expected outcome and not an achievement — at this length a collision would mean something was badly broken — but it is worth confirming rather than assuming.

Why there is no honest crack-time promise

Guessing speed depends on the attack. Online sign-in attempts may be constrained by rate limits and monitoring; offline attacks depend on the service's password hashing method, its parameters, the stolen data, and the attacker's hardware. Phishing, malware, an exposed recovery channel, or a compromised service can bypass brute force entirely. No generated password is a guarantee against those failures.

Store passwords instead of memorizing them

A reputable password manager can generate and store unique credentials behind one well-protected vault. Protect that vault with a strong master passphrase and multifactor authentication. Where a service offers passkeys, they can reduce phishing risk because authentication is bound to the legitimate site rather than a reusable secret.

When to change a password

Change it promptly when the service reports a breach, the password was reused, it was shared or exposed, or you suspect account compromise. Routine changes on an arbitrary calendar can encourage weaker patterns; follow the service's incident guidance and current organizational policy instead.

Practical checklist

  • Generate a unique password for each account.
  • Use a password manager and protect its recovery options.
  • Enable phishing-resistant MFA or passkeys when available.
  • Keep the browser, operating system, and extensions updated.
  • Never paste a generated secret into an untrusted page or message.

How this guide was reviewed

We inspected the generator's use of crypto.getRandomValues, compared its options with current NIST and OWASP guidance, and removed claims that any password is uncrackable.

Last technical review: .

Sources reviewed

Related Articles

Learn more with related in-depth guides and tutorials.