Date Range Splitter

The split is computed in milliseconds and each boundary is then snapped to local midnight, so some chunks may run a day longer or shorter than others. The rounding is spread across the boundaries rather than dumped on the last chunk, keeping the intervals as balanced as whole-day output allows.

Boundaries are exactly the cut-points you'd feed into a BETWEEN clause, a partitioning function, or a loop over date windows — no start/end pairing needed. Intervals is better when a human reads the schedule and wants each chunk's length in days at a glance.

The rows follow the half-open convention [start, end): the shared boundary belongs to the later chunk, so no day is counted twice and none falls into a gap. If your report needs inclusive per-chunk end dates instead, subtract one day from each end column after copying.

All dates are interpreted at midnight in your browser's local timezone, and every computed boundary is snapped back to local midnight, so the output is always clean whole dates — never partial days or odd times. A DST transition inside the range shifts the raw millisecond cut by an hour; after snapping, that can at most move a boundary landing near midnight onto the neighbouring day, the same whole-day rounding described above.

No — it divides the span into equal-length slices by elapsed time. Splitting a year into 4 gives ~91-day chunks, which only approximate calendar quarters (Jan–Mar is 90 days, Jul–Sep is 92). For strict calendar periods, list the month or quarter start dates yourself; use this tool when the chunks must be equal.