C numeric limits interface

See also std::numeric_limits interface.

Characteristics of integer types

Defined in header `<climits>`
Characteristics of core language integer types
BOOL_WIDTH (C++26)	bit width of bool (macro constant)
CHAR_BIT	bit width of byte (macro constant)
MB_LEN_MAX	maximum number of bytes in a multibyte character (macro constant)
CHAR_WIDTH (C++26)	bit width of char (macro constant)
CHAR_MIN	minimum value of char (macro constant)
CHAR_MAX	maximum value of char (macro constant)
SCHAR_WIDTHSHRT_WIDTHINT_WIDTHLONG_WIDTHLLONG_WIDTH (C++26)(C++26)(C++26)(C++26)(C++26)	bit width of signedchar, short, int, long, and longlong respectively (macro constant)
SCHAR_MINSHRT_MININT_MINLONG_MINLLONG_MIN (C++11)	minimum value of signedchar, short, int, long and longlong respectively (macro constant)
SCHAR_MAXSHRT_MAXINT_MAXLONG_MAXLLONG_MAX (C++11)	maximum value of signedchar, short, int, long and longlong respectively (macro constant)
UCHAR_WIDTHUSHRT_WIDTHUINT_WIDTHULONG_WIDTHULLONG_WIDTH (C++26)(C++26)(C++26)(C++26)(C++26)	bit width of unsignedchar, unsignedshort, unsignedint, unsignedlong, and unsignedlonglong respectively (macro constant)
UCHAR_MAXUSHRT_MAXUINT_MAXULONG_MAXULLONG_MAX (C++11)	maximum value of unsignedchar, unsignedshort, unsignedint, unsignedlong and unsignedlonglong respectively (macro constant)
Defined in header `<cwchar>`
Defined in header `<cstdint>`
WCHAR_WIDTH (C++26)	bit width of wchar_t (macro constant)
WCHAR_MIN (C++11)	minimum value of wchar_t (macro constant)
WCHAR_MAX (C++11)	maximum value of wchar_t (macro constant)
Characteristics of library type aliases
Defined in header `<cstdint>`
PTRDIFF_WIDTH (C++26)	bit width of `std::ptrdiff_t` (macro constant)
PTRDIFF_MIN (C++11)	minimum value of `std::ptrdiff_t` (macro constant)
PTRDIFF_MAX (C++11)	maximum value of `std::ptrdiff_t` (macro constant)
SIZE_WIDTH (C++26)	bit width of `std::size_t` (macro constant)
SIZE_MAX (C++11)	maximum value of `std::size_t` (macro constant)
SIG_ATOMIC_WIDTH (C++26)	bit width of `std::sig_atomic_t` (macro constant)
SIG_ATOMIC_MIN (C++11)	minimum value of `std::sig_atomic_t` (macro constant)
SIG_ATOMIC_MAX (C++11)	maximum value of `std::sig_atomic_t` (macro constant)
WINT_WIDTH (C++26)	bit width of `std::wint_t` (macro constant)
WINT_MIN (C++11)	minimum value of `std::wint_t` (macro constant)
WINT_MAX (C++11)	maximum value of `std::wint_t` (macro constant)

Value requirements

BOOL_WIDTH always expands to 1.

(since C++26)

Each of the *_WIDTH macros (other than BOOL_WIDTH),(since C++26) CHAR_BIT and MB_LEN_MAX expands to an implementation-defined value. The allowed minimum values are listed below, where each group of macros separated by / always expand to the same value:

at least 1: MB_LEN_MAX
at least 8: CHAR_BIT / CHAR_WIDTH / SCHAR_WIDTH / UCHAR_WIDTH, WCHAR_WIDTH, SIG_ATOMIC_WIDTH
at least 16: SHRT_WIDTH / USHRT_WIDTH, INT_WIDTH / UINT_WIDTH, PTRDIFF_WIDTH, SIZE_WIDTH, WINT_WIDTH
at least 32: LONG_WIDTH / ULONG_WIDTH
at least 64: LLONG_WIDTH / ULLONG_WIDTH

Notes

The types of these constants, other than the *_WIDTH macros, (since C++26)CHAR_BIT and MB_LEN_MAX, are required to match the results of the integral promotions as applied to objects of the types they describe: CHAR_MAX may have type int or unsignedint, but never char. Similarly USHRT_MAX may not be of an unsigned type: its type may be int.

A freestanding implementation may lack std::sig_atomic_t and/or std::wint_t typedef names, in which case the SIG_ATOMIC_* and/or WINT_* macros are correspondingly absent.

Example

#include <climits>
#include <cstdint>
#include <iomanip>
#include <iostream>
#include <locale>

int main()
{
    constexpr int w = 14;
    (std::cout << std::left).imbue(std::locale("en_US.UTF-8"));
#   define COUT(x) std::cout << std::setw(w) << #x << " = " << x << '\n'
    
    COUT( BOOL_WIDTH     );
    COUT( CHAR_BIT       );
    COUT( CHAR_MAX       );
    COUT( CHAR_MIN       );
    COUT( CHAR_WIDTH     );
    COUT( INT_MAX        );
    COUT( INT_MIN        );
    COUT( INT_WIDTH      );
    COUT( LLONG_MAX      );
    COUT( LLONG_MIN      );
    COUT( LLONG_WIDTH    );
    COUT( LONG_MAX       );
    COUT( LONG_MIN       );
    COUT( LONG_WIDTH     );
    COUT( MB_LEN_MAX     );
    COUT( PTRDIFF_MAX    );
    COUT( PTRDIFF_MIN    );
    COUT( SCHAR_MAX      );
    COUT( SCHAR_MIN      );
    COUT( SCHAR_WIDTH    );
    COUT( SHRT_MAX       );
    COUT( SHRT_MIN       );
    COUT( SHRT_WIDTH     );
    COUT( SIG_ATOMIC_MAX );
    COUT( SIG_ATOMIC_MIN );
    COUT( SIZE_MAX       );
    COUT( UCHAR_MAX      );
    COUT( UCHAR_WIDTH    );
    COUT( UINT_MAX       );
    COUT( UINT_WIDTH     );
    COUT( ULLONG_MAX     );
    COUT( ULLONG_WIDTH   );
    COUT( ULONG_MAX      );
    COUT( ULONG_WIDTH    );
    COUT( USHRT_MAX      );
    COUT( USHRT_WIDTH    );
    COUT( WCHAR_MAX      );
    COUT( WCHAR_MIN      );
    COUT( WINT_MAX       );
    COUT( WINT_MIN       );
}

Possible output:

BOOL_WIDTH     = 1
CHAR_BIT       = 8
CHAR_MAX       = 127
CHAR_MIN       = -128
CHAR_WIDTH     = 8
INT_MAX        = 2,147,483,647
INT_MIN        = -2,147,483,648
INT_WIDTH      = 32
LLONG_MAX      = 9,223,372,036,854,775,807
LLONG_MIN      = -9,223,372,036,854,775,808
LLONG_WIDTH    = 64
LONG_MAX       = 9,223,372,036,854,775,807
LONG_MIN       = -9,223,372,036,854,775,808
LONG_WIDTH     = 64
MB_LEN_MAX     = 16
PTRDIFF_MAX    = 9,223,372,036,854,775,807
PTRDIFF_MIN    = -9,223,372,036,854,775,808
SCHAR_MAX      = 127
SCHAR_MIN      = -128
SCHAR_WIDTH    = 8
SHRT_MAX       = 32,767
SHRT_MIN       = -32,768
SHRT_WIDTH     = 16
SIG_ATOMIC_MAX = 2,147,483,647
SIG_ATOMIC_MIN = -2,147,483,648
SIZE_MAX       = 18,446,744,073,709,551,615
UCHAR_MAX      = 255
UCHAR_WIDTH    = 8
UINT_MAX       = 4,294,967,295
UINT_WIDTH     = 32
ULLONG_MAX     = 18,446,744,073,709,551,615
ULLONG_WIDTH   = 64
ULONG_MAX      = 18,446,744,073,709,551,615
ULONG_WIDTH    = 64
USHRT_MAX      = 65,535
USHRT_WIDTH    = 16
WCHAR_MAX      = 2,147,483,647
WCHAR_MIN      = -2,147,483,648
WINT_MAX       = 4,294,967,295
WINT_MIN       = 0

Characteristics and special values of floating-point types

Defined in header `<cfloat>`
Characteristics of floating-point types
FLT_ROUNDS	default rounding mode of floating-point arithmetic (macro constant)
FLT_EVAL_METHOD (C++11)	specifies in what precision all arithmetic operations are done (macro constant)

FLT_HAS_SUBNORMDBL_HAS_SUBNORMLDBL_HAS_SUBNORM

(C++17)(deprecated in C++26)

specifies whether the type supports subnormal (denormal) numbers:
-1 – indeterminable, 0 – absent, 1 – present
(macro constant)

FLT_RADIX	the radix (integer base) used by the representation of all three floating-point types (macro constant)
FLT_MANT_DIGDBL_MANT_DIGLDBL_MANT_DIG	number of base `FLT_RADIX` digits that can be represented without losing precision for float, double and longdouble respectively (macro constant)
FLT_DECIMAL_DIGDBL_DECIMAL_DIGLDBL_DECIMAL_DIG (C++17)	conversion from float/double/longdouble to decimal with at least `FLT_DECIMAL_DIG`/`DBL_DECIMAL_DIG`/`LDBL_DECIMAL_DIG` digits and back is the identity conversion: this is the decimal precision required to serialize/deserialize a floating-point value (macro constant)

DECIMAL_DIG

(C++11)(deprecated in C++26)

conversion from longdouble to decimal with at least DECIMAL_DIG digits and back to longdouble is the identity conversion: this is the decimal precision required to serialize/deserialize a longdouble
(macro constant)

Defined in header `<cmath>` (deprecated in C++26)
FLT_DIGDBL_DIGLDBL_DIG	number of decimal digits that are guaranteed to be preserved in text → float/double/longdouble → text roundtrip without change due to rounding or overflow (macro constant)
FLT_MIN_EXPDBL_MIN_EXPLDBL_MIN_EXP	minimum negative integer such that `FLT_RADIX` raised by power one less than that integer is a normalized float, double and longdouble respectively (macro constant)
FLT_MIN_10_EXPDBL_MIN_10_EXPLDBL_MIN_10_EXP	minimum negative integer such that `10` raised to that power is a normalized float, double and longdouble respectively (macro constant)
FLT_MAX_EXPDBL_MAX_EXPLDBL_MAX_EXP	maximum positive integer such that `FLT_RADIX` raised by power one less than that integer is a representable finite float, double and long double respectively (macro constant)
FLT_MAX_10_EXPDBL_MAX_10_EXPLDBL_MAX_10_EXP	maximum positive integer such that `10` raised to that power is a representable finite float, double and long double respectively (macro constant)
FLT_MAXDBL_MAXLDBL_MAX	maximum finite value of float, double and long double respectively (macro constant)
FLT_EPSILONDBL_EPSILONLDBL_EPSILON	difference between `1.0` and the next representable value for float, double and long double respectively (macro constant)
FLT_MINDBL_MINLDBL_MIN	minimum normalized positive value of float, double and long double respectively (macro constant)
FLT_TRUE_MINDBL_TRUE_MINLDBL_TRUE_MIN (C++17)	minimum positive value of float, double and long double respectively (macro constant)
Special values of floating-point types
Defined in header `<cfloat>` (since C++26)
INFINITY (C++11)	evaluates to positive infinity (if exists) (macro constant)
NAN (C++11)	evaluates to a quiet NaN of type float (if exists) (macro constant)
Defined in header `<cfloat>`
FLT_SNANDBL_SNANLDBL_SNAN (C++26)	evaluates to a signaling NaN of float, double and long double respectively (if exists) (macro constant)

Example

#include <cfloat>
#include <iomanip>
#include <iostream>

int main()
{
    int w = 16;
    std::cout << std::left; // std::cout << std::setprecision(53);
#   define COUT(x) std::cout << std::setw(w) << #x << " = " << x << '\n'
    
    COUT( FLT_RADIX        );
    COUT( DECIMAL_DIG      );
    COUT( FLT_DECIMAL_DIG  );
    COUT( DBL_DECIMAL_DIG  );
    COUT( LDBL_DECIMAL_DIG );
    COUT( FLT_MIN          );
    COUT( DBL_MIN          );
    COUT( LDBL_MIN         );
    COUT( FLT_TRUE_MIN     );
    COUT( DBL_TRUE_MIN     );
    COUT( LDBL_TRUE_MIN    );
    COUT( FLT_MAX          );
    COUT( DBL_MAX          );
    COUT( LDBL_MAX         );
    COUT( FLT_EPSILON      );
    COUT( DBL_EPSILON      );
    COUT( LDBL_EPSILON     );
    COUT( FLT_DIG          );
    COUT( DBL_DIG          );
    COUT( LDBL_DIG         );
    COUT( FLT_MANT_DIG     );
    COUT( DBL_MANT_DIG     );
    COUT( LDBL_MANT_DIG    );
    COUT( FLT_MIN_EXP      );
    COUT( DBL_MIN_EXP      );
    COUT( LDBL_MIN_EXP     );
    COUT( FLT_MIN_10_EXP   );
    COUT( DBL_MIN_10_EXP   );
    COUT( LDBL_MIN_10_EXP  );
    COUT( FLT_MAX_EXP      );
    COUT( DBL_MAX_EXP      );
    COUT( LDBL_MAX_EXP     );
    COUT( FLT_MAX_10_EXP   );
    COUT( DBL_MAX_10_EXP   );
    COUT( LDBL_MAX_10_EXP  );
    COUT( FLT_ROUNDS       );
    COUT( FLT_EVAL_METHOD  );
    COUT( FLT_HAS_SUBNORM  );
    COUT( DBL_HAS_SUBNORM  );
    COUT( LDBL_HAS_SUBNORM );
}

Possible output:

FLT_RADIX        = 2
DECIMAL_DIG      = 21
FLT_DECIMAL_DIG  = 9
DBL_DECIMAL_DIG  = 17
LDBL_DECIMAL_DIG = 21
FLT_MIN          = 1.17549e-38
DBL_MIN          = 2.22507e-308
LDBL_MIN         = 3.3621e-4932
FLT_TRUE_MIN     = 1.4013e-45
DBL_TRUE_MIN     = 4.94066e-324
LDBL_TRUE_MIN    = 3.6452e-4951
FLT_MAX          = 3.40282e+38
DBL_MAX          = 1.79769e+308
LDBL_MAX         = 1.18973e+4932
FLT_EPSILON      = 1.19209e-07
DBL_EPSILON      = 2.22045e-16
LDBL_EPSILON     = 1.0842e-19
FLT_DIG          = 6
DBL_DIG          = 15
LDBL_DIG         = 18
FLT_MANT_DIG     = 24
DBL_MANT_DIG     = 53
LDBL_MANT_DIG    = 64
FLT_MIN_EXP      = -125
DBL_MIN_EXP      = -1021
LDBL_MIN_EXP     = -16381
FLT_MIN_10_EXP   = -37
DBL_MIN_10_EXP   = -307
LDBL_MIN_10_EXP  = -4931
FLT_MAX_EXP      = 128
DBL_MAX_EXP      = 1024
LDBL_MAX_EXP     = 16384
FLT_MAX_10_EXP   = 38
DBL_MAX_10_EXP   = 308
LDBL_MAX_10_EXP  = 4932
FLT_ROUNDS       = 1
FLT_EVAL_METHOD  = 0
FLT_HAS_SUBNORM  = 1
DBL_HAS_SUBNORM  = 1
LDBL_HAS_SUBNORM = 1

Defect reports

The following behavior-changing defect reports were applied retroactively to previously published C++ standards.

DR	Applied to	Behavior as published	Correct behavior
LWG 416	C++98	it was unclear whether the types of the macros in `<climits>` are guaranteed to match the type to which they refer (C++ refers to C, and C says no)	clarified as not guaranteed

C numeric limits interface

Characteristics of integer types

Characteristics of core language integer types

Characteristics of library type aliases

Value requirements

Notes

Example

Characteristics and special values of floating-point types

Characteristics of floating-point types

Special values of floating-point types

Example

Defect reports

See also