Scientific Notation Converter

Convert exact decimals to scientific, engineering and plain notation with explicit rounding options.

Convert up to 1,000 decimal rows into scientific, engineering and ordinary notation. Coefficient digits are handled as text, so long values are never rounded through a JavaScript floating-point number.

Conversion and downloads stay in this tab. Inputs are not evaluated, uploaded or stored. CSV protects a leading plus in the original input column as text; JSON and TXT retain the exact accepted lexeme.

This tool processes its input contents in your browser instead of sending them to a processing service. Page, asset and fixed usage-measurement requests are described on the Privacy page.

Converts up to 1,000 exact decimal lexemes into scientific, engineering and bounded plain notation with optional 2–16-digit round-half-to-even output.

How to use it

1Enter one signed decimal or e-notation value per nonblank line.
2Keep exact coefficient digits or choose 2–16 significant figures.
3Convert and review the first 20 rows with their original source-line numbers.
4Download all rows as CSV, JSON or TXT.

Example

Input: 12345
Output: Scientific: 1.2345e4 · Engineering: 12.345e3 · Plain: 12345

This exact example, long coefficients, negative zero, exponent bounds and rounding carry are covered by the focused helper check.

What happens to your data

Decimal inputs are parsed as text and remain in this tab without upload or browser storage. Fixed anonymous action events can contain the action and row count, never the entered numbers. Downloads are created only after an explicit action.

Last updated November 2026

Convert one decimal or up to 1,000 decimal rows into scientific notation, engineering notation and ordinary decimal notation. Exact mode retains the entered coefficient digits, including trailing zeros, without first turning the value into a JavaScript floating-point number. Rounded mode applies one explicit round-half-to-even step and pads every nonzero result to a selected 2–16 significant figures.

The converter accepts decimal text and e notation only. It does not evaluate formulas, arithmetic, fractions, commas, units, NaN or Infinity. This keeps the result about notation and decimal digits rather than turning the page into an expression evaluator or duplicating the separate significant-figures arithmetic tool.

How it works

Parsing separates the sign, integer digits, fraction digits and optional base-10 exponent. Leading zeros determine the normalized exponent but do not pass through Number, parseFloat or binary floating point. For a nonzero value, scientific notation places one digit before the decimal point. Engineering notation chooses the greatest exponent divisible by three that is no greater than the scientific exponent, leaving one, two or three integer digits in the coefficient. Thus 12345 becomes 1.2345e4 in scientific notation and 12.345e3 in engineering notation.

Exact mode keeps every coefficient digit after leading zeros are removed. That means 0.0012300 becomes 1.2300e-3, 1.2300e-3 and 0.0012300. It preserves an entered negative zero and its decimal scale in plain output where representable. It does not claim that an entered exponent is itself significant metadata: equivalent values can normalize to the same output exponent.

Rounded mode keeps the selected leading digits, inspects all discarded digits once and uses round half to even. An exact halfway case increments only when the last kept digit is odd; any nonzero digit after the first discarded 5 makes the value greater than halfway. Carry can increase the exponent: 999.5 rounded to three significant figures becomes 1.00e3, 1.00e3 in engineering notation and 1000. Scientific and engineering strings retain the selected significant-figure padding; ordinary integer notation cannot always communicate those trailing significant zeros without an external convention.

Each coefficient is limited to 1,000 entered digits, each lexeme to 2,048 characters, and the exponent magnitude to 10,000. Plain notation is limited to 10,000 characters per value and all generated text to four million characters, so an exponent cannot produce an unbounded run of zeros. A failed row identifies its source line and produces no stale download. Blank lines are ignored but their original line numbers are retained. CSV, JSON and TXT exports contain every converted row; the preview shows at most 20. CSV prefixes a leading plus in the original-input column with an apostrophe so spreadsheet software does not interpret it as a formula. JSON and TXT retain the exact accepted input lexeme. Inputs and results stay in this tab without uploads or storage. Fixed anonymous action events can report the action and row count but never the entered numbers.

Worked example: Normalize an integer without binary floating-point conversion

Input: 12345
Result: Scientific: 1.2345e4 · Engineering: 12.345e3 · Plain: 12345

This exact example, long coefficients, negative zero, exponent bounds and rounding carry are covered by the focused helper check.

Common use cases

Converting measurements into normalized scientific notation without losing long decimal digits
Formatting powers of ten in engineering groups of three for technical notes
Expanding bounded e notation into ordinary decimal text
Batch-exporting exact conversions with original source-line provenance
Demonstrating round-half-to-even carry at a chosen significant-figure count

Frequently asked questions

What is the difference between scientific and engineering notation?＋

Scientific notation uses exactly one nonzero digit before the decimal point. Engineering notation uses an exponent divisible by three and therefore has one, two or three integer digits in the coefficient. For example, 12345 is 1.2345e4 scientifically and 12.345e3 in engineering notation.

Will a long decimal be rounded by JavaScript?＋

Not by this converter. Accepted coefficient digits remain strings through parsing, normalization and rounding. Small bounded settings such as line numbers and exponents use ordinary integers, but the decimal value itself is never converted to a JavaScript floating-point number.

What does exact mode preserve?＋

It preserves the accepted coefficient digits after irrelevant leading zeros are removed, including trailing zeros, and preserves a minus sign on zero. It normalizes notation, so it does not preserve the original placement of the decimal point, the letter case of E or the entered exponent spelling.

How are halfway values rounded?＋

Rounded mode uses round half to even. At an exact tie, the last kept digit stays when it is even and increments when it is odd. Any later nonzero discarded digit makes the value greater than halfway and therefore rounds up.

Why can ordinary notation be refused for a valid exponent?＋

Writing a value such as 1e-10000 in plain form requires thousands of zeros. Per-value and batch output limits prevent a small input from producing unbounded text. Use scientific or engineering notation for magnitudes whose plain representation exceeds the displayed limit.

Can I paste calculations such as 6.02e23 / 2?＋

No. The page accepts one decimal lexeme per nonblank line and performs no expression evaluation. Calculate the value separately, then paste the decimal or e-notation result you want to reformat.

Do the downloads contain my exact inputs?＋

JSON and TXT retain each accepted trimmed lexeme. CSV also retains it, except that a leading plus is prefixed with an apostrophe as a spreadsheet-safety measure. Converted values are generated locally and are not uploaded or stored by the tool.