How to convert a string to an int in Python

1The int() built-inRecommended

int(text) parses a decimal integer and hands back a real int you can do arithmetic with. Leading and trailing whitespace is stripped for you, and a leading - or + is honoured.

text = "42"
number = int(text)

print(number, type(number))
print(number + 8)
print(int("  -7  "))

Prints 42 <class 'int'>, then 50, then -7. Note that number + 8 gives 50, not "428" — that concatenation is exactly the bug the conversion exists to prevent.

2Handle bad input with try/except

Anything int() can't parse raises ValueError, and a None raises TypeError. Since user input, CSV cells and query strings are never trustworthy, the real-world shape is a small helper that catches both and returns a default. The tempting alternative — check first with .isdigit() — is printed alongside so you can see it disagree.

def to_int(value, default=None):
    try:
        return int(value)
    except (ValueError, TypeError):
        return default


# The "check first" alternative — weaker than it looks.
def looks_like_int(text):
    return text.strip().lstrip("-").isdigit()


for raw in ["42", "  -7  ", "+5", "3.5", "abc", "²"]:
    print(f"{raw!r:>8}  int -> {to_int(raw)!r:>6}   isdigit check -> {looks_like_int(raw)}")

The check is wrong in both directions. It rejects "+5", which int() happily parses as 5, because lstrip("-") knows nothing about a leading plus. And it accepts "²" — str.isdigit() is true for superscripts and other Unicode digit characters that int() flatly refuses. Catching the exception is the only check that agrees with the parser, because it is the parser.

3Strings with a decimal point

int("3.5") does not round — it raises ValueError: invalid literal for int() with base 10: '3.5'. Go through float() first. Just know what you're buying: int() truncates toward zero, while round() rounds to nearest and breaks ties toward the even number.

price = "3.5"

try:
    int(price)
except ValueError as err:
    print("int('3.5') ->", err)

print(int(float(price)))       # truncates toward zero
print(int(float("-3.5")))
print(round(float(price)))     # rounds — and ties go to even
print(round(float("2.5")))

Output: the ValueError message, then 3, -3, 4, 2. Truncation is symmetric around zero — -3.5 becomes -3, not -4. And round(2.5) is 2, not 3: Python uses banker's rounding, so exact .5 ties land on the even integer.

4Binary, hex, and other bases

int() takes a second argument: the base. Pass 2 for binary, 8 for octal, 16 for hex. The matching 0b/0o/0x prefix is optional but allowed. Pass 0 and Python reads the prefix and picks the base itself.

print(int("1010", 2))
print(int("ff", 16))
print(int("0xff", 16))
print(int("755", 8))

# base 0 means "read the prefix and work it out"
print(int("0b1010", 0), int("0o755", 0), int("0xff", 0), int("42", 0))

try:
    int("0755", 0)
except ValueError as err:
    print("int('0755', 0) ->", err)

Prints 10, 255, 255, 493, then 10 493 255 42. The last line is the catch: with base 0 a string must be a valid Python literal, and "0755" — a C-style octal with no o — isn't one, so it raises invalid literal for int() with base 0: '0755'. If you know the base, say so explicitly.

5Which should you use?

6Common variation: a whole list of strings

For a clean list, map(int, rows) is the shortest thing that works — and it's lazy, so sum(map(int, rows)) never builds the intermediate list. For messy data, map a forgiving helper instead and filter out the misses.

rows = ["1", "2", "3", "4"]

print(list(map(int, rows)))
print(sum(map(int, rows)))


def to_int(text):
    try:
        return int(text)
    except ValueError:
        return None


messy = ["10", "x", "20", "", "30"]
numbers = [n for n in map(to_int, messy) if n is not None]
print(numbers, sum(numbers))

Prints [1, 2, 3, 4], 10, then [10, 20, 30] 60. Filter on is not None rather than truthiness — "0" converts to 0, which is falsy, and a plain if n would silently drop it.