How to use it
- Choose A number and type a value, or choose Hex bits and paste 8 or 16 hex digits.
- Read the float32 and float64 rows: hex, bit groups, sign, exponent, mantissa and the exact stored value.
- Copy any row with its copy button where one is offered.
How it works
The number is written into a 32-bit and a 64-bit IEEE 754 binary format using the browser’s own floating-point conversion, and the bits are read back. The exact value is calculated from the integer significand and the power of two with big-integer arithmetic, so every digit shown is real.
Example: 0.1 in float32 is 0x3DCCCCCD, with bits 0 01111011 10011001100110011001101. In float64 it is 0x3FB999999999999A.
Tips and common mistakes
- -0 and 0 have different bits (the sign bit) but compare equal.
- Integers above 2^24 (float32) or 2^53 (float64) cannot all be stored exactly.
- Hex is shown as big-endian text; a file or network dump may list the bytes in little-endian order.