Introduction

Getting the current time in Python is easy, but the right function depends on what you actually need. Sometimes you want a local wall-clock time for display, sometimes a timezone-aware UTC timestamp for storage, and sometimes just epoch seconds for measurement.

The biggest mistake is treating all of those as the same thing. Python gives you several time APIs because they solve different problems.

Use datetime.now() for a Local Date and Time

If you want the current local date and time, the most direct option is datetime.now().

1from datetime import datetime
2
3current = datetime.now()
4print(current)

This returns a datetime object representing the current local time according to the machine running the code. It is convenient for quick scripts, logs, and human-readable output.

You can format it with strftime:

1from datetime import datetime
2
3current = datetime.now()
4formatted = current.strftime("%Y-%m-%d %H:%M:%S")
5print(formatted)

Prefer Timezone-Aware UTC for Storage

For many applications, UTC is a better default than local time. It is unambiguous and easier to compare across systems.

1from datetime import datetime, timezone
2
3current_utc = datetime.now(timezone.utc)
4print(current_utc)

This returns a timezone-aware datetime. That is usually preferable to a naive local datetime when writing timestamps to databases, APIs, or logs that may be read in multiple time zones.

If You Only Need the Time Portion

Sometimes you want just the current time of day.

1from datetime import datetime
2
3now = datetime.now()
4current_time_only = now.time()
5print(current_time_only)

This gives you a time object, which is useful for display or scheduling logic that does not need the full date.

Use time.time() for Epoch Seconds

If you need a Unix timestamp, use time.time().

1import time
2
3seconds_since_epoch = time.time()
4print(seconds_since_epoch)

This returns the number of seconds since the Unix epoch as a floating-point value. It is useful for simple timing, interoperability, or compact numeric storage.

Do not confuse this with a formatted human time. It is a machine-oriented representation.

Measuring Elapsed Time Is a Different Task

If your goal is performance measurement, time.time() is not always the best tool. Python provides timers intended for elapsed time measurement.

1import time
2
3start = time.perf_counter()
4time.sleep(0.5)
5end = time.perf_counter()
6
7print("Elapsed:", end - start)

This is better for benchmarking because it is designed for measuring durations, not just reporting the current wall-clock time.

Converting to Another Time Zone

Once you have a timezone-aware datetime, converting it is straightforward.

1from datetime import datetime, timezone
2from zoneinfo import ZoneInfo
3
4utc_now = datetime.now(timezone.utc)
5toronto_now = utc_now.astimezone(ZoneInfo("America/Toronto"))
6print(toronto_now)

Using zoneinfo from the standard library is the modern built-in way to work with time zones in current Python versions.

Which One Should You Choose?

A useful rule is:

• use datetime.now() for local human-facing time
• use datetime.now(timezone.utc) for stored or transmitted timestamps
• use time.time() for epoch seconds
• use time.perf_counter() for measuring elapsed time

When you choose the function based on the job, the code becomes clearer and fewer timezone bugs slip in later.

Common Pitfalls

• Using local naive datetimes for database or API timestamps that really should be UTC.
• Treating time.time() as if it were a readable local time string.
• Using wall-clock time APIs to measure code performance instead of perf_counter().
• Forgetting that a datetime without timezone information can become ambiguous when shared across systems.
• Formatting time too early and then losing the ability to do date arithmetic cleanly.

Summary

• 'datetime.now() gives you the current local date and time.'
• 'datetime.now(timezone.utc) is usually the safer choice for storage and APIs.'
• 'time.time() returns epoch seconds, not a human-friendly formatted timestamp.'
• 'time.perf_counter() is the right tool for measuring elapsed time.'
• Pick the time API based on whether you need display, storage, interoperability, or benchmarking.