Convert numbers in any base up to 62, where digits are case-sensitive: 0-9, then A-Z for 10-35, then a-z for 36-61. Base 62 is the alphabet behind short URLs and compact IDs.
Show calculation steps
Processed privately in your browser — nothing you paste is uploaded, logged or stored.
The case-sensitive frontier
Past base 36 the digit alphabet runs out of case-insensitive symbols, so base 62 recruits both cases: 0–9, then A–Z for 10–35, then a–z for 36–61. The letter A and the letter a are different digits here — values 10 and 36 — which makes this converter case-sensitive throughout, unlike its base-36 sibling.
Base 62 matters because its alphabet is exactly the URL-safe alphanumerics: no escaping, no padding, maximum density. Six characters index 56.8 billion values; eight reach 218 trillion.
Worked example
Input: 9007199254740991 (decimal, i.e. 2^53−1) → base 62 Result: fFgnDxSe7 (9 characters instead of 16 digits) Round trip: fFgnDxSe7 → 9007199254740991 ✓ exact
Where base 62 shows up
URL shorteners encode database row IDs as short slugs; video platforms and paste sites generate base-62-style tokens; ticketing and referral systems compress numeric IDs for humans to retype. Decoding such an identifier back to its number — or predicting how long an ID space lasts — is this page’s daily work.
A warning about case
Because case carries value, transports that fold case (some legacy systems, case-insensitive filesystems, shouting users) corrupt base-62 data irreversibly: FFF and fff differ by hundreds of thousands. If your channel cannot guarantee case, step down to base 36 — that is precisely the trade-off between the two alphabets.
Exactness and privacy
Conversions are BigInt-exact at any length — the 2⁵³−1 example above is the ceiling where floating-point tools begin to lie, not where this one does. As with every tool in this cluster, the computation is local; identifiers you paste stay on your device.
How to use the Base 2 to 62 Converter
- Enter the number and select its base (2-62).
- Select the target base.
- Click "Convert base". Above base 36 the digit alphabet is case-sensitive: 0-9, then A-Z for 10-35, then a-z for 36-61.
- Copy the result – base-62 output is URL-safe with no padding characters.
Frequently asked questions
Why does letter case suddenly matter here?
Because beyond 36 symbols the alphabet must distinguish A from a: in base 62, "A" is 10 but "a" is 36. This converter is therefore case-sensitive everywhere, which is the key behavioural difference from the 2-36 converter.
What is base 62 used for in practice?
Compact identifiers: URL shorteners, YouTube-style video IDs and database short-keys, because 0-9A-Za-z is the largest alphabet that needs no escaping in URLs. Six base-62 characters cover 56.8 billion combinations.
How much shorter is base 62 than decimal?
Roughly 45 percent – the length ratio is log(10)/log(62) ≈ 0.558. The sample value 9007199254740991 is 16 decimal digits but only 9 characters in base 62 (fFgnDxSe7).
Is this the same as Base64 encoding?
No, and the distinction matters. Base 62 here is a NUMBER base – positional notation for integer values. Base64 is a byte-stream encoding with 64 symbols including + / and = padding, applied to arbitrary data. Converting the number 100 to base 62 and Base64-encoding the string "100" produce unrelated results.
Can I convert between two high bases directly, say base 45 to base 58?
Yes – any pair from 2 to 62 works in one pass. Internally the value crosses an exact decimal bridge, so no precision is lost regardless of the pair.
Does it support negative numbers?
Yes, with a leading minus sign. Note that real-world base-62 ID systems never use signs – if you are decoding a short URL ID, the value should be non-negative.
How large can the input be?
Unbounded in practice: BigInt arithmetic keeps a 100-character base-62 string exact, which is far beyond any 64-bit converter.
