Words to Number Converter

About this tool

The Words to Number Converter parses English number words back into digits. Type or paste text like 'forty-two thousand three hundred and twelve' and the tool returns 42,312, both with thousands separators and as a raw number you can copy. It handles hundreds, thousands, millions, billions and trillions, hyphenated or spaced compounds ('twenty-one' or 'twenty one'), the connector words 'and', 'a' and 'the', the ordinals first to twelfth (including compounds such as 'twenty-first'), hundredth and thousandth, and negative values introduced by 'negative' or 'minus'. Decimals work with 'point' (or 'dot') followed by single digit words, so 'three point one four' gives 3.14. Very large values are computed exactly with big integers rather than rounded. Words the parser does not recognise are skipped silently, so check the result when pasting longer text.

Why use it

• Parses everyday English number words from single digits to trillions, including decimals.

• Handles negative values and common ordinals such as 'twenty-first'.

• Useful for processing voice-to-text transcriptions with number words.

• 100% browser-based — no API call needed.

• Accepts British-style 'and' and either hyphens or spaces in compounds.

• Saves time when migrating legacy records that mix words and digits.

How to use

1. Type or paste English number words into the input field.
2. The numeric equivalent appears instantly, with and without thousands separators.
3. Click 'Copy number' to copy the raw digits.
4. Enter one number at a time — every recognised number word in the box is combined into a single value.
5. Remove non-number words first if the result looks wrong, since unknown words are skipped silently.
6. Use the result alongside the Number to Words tool to verify round-trip accuracy.

When it helps

• Converting transcribed speech or dictation containing number words to digits.
• Parsing number words in imported documents before spreadsheet processing.
• Testing or building NLP systems that handle number word extraction.
• Checking that written check amounts correspond to the expected numeric value.
• Preparing data for spreadsheet import where amounts were transcribed in words.
• Verifying that voice-to-text output captures dictated numbers correctly.

Examples

Tips

• Hyphens and 'and' are treated as connectors — both 'twenty-one' and 'twenty one' parse to 21.
• Read decimals digit by digit after 'point' — 'two point five zero' gives 2.50, while 'point twenty-five' is not understood.
• When parsing transcribed speech, pre-clean the text by removing filler words ('um', 'like') for cleaner output.
• The parser is forgiving with case — uppercase, lowercase, and mixed all work the same.
• For negative values use 'negative' or 'minus' as a prefix — both are recognised.
• If your text has multiple numbers, parse them one at a time — all number words in the box are combined into one result.
• Ordinals (first, second, twenty-first) are converted to their cardinal equivalent (1, 2, 21) for downstream calculations.

Frequently Asked Questions

Glossary

Cardinal number

A number used for counting: one, two, three. Cardinals indicate quantity.

Ordinal number

A number indicating position or rank: first, second, third. The parser converts these to their cardinal equivalents.

Compound number

A number formed by combining smaller numbers, like 'twenty-three' (20 + 3) or 'three hundred forty-five' (300 + 45).

Place-value words

Words like 'hundred', 'thousand', 'million' that multiply the preceding number rather than adding to it.

Connector word

Words like 'and' that join number components without contributing a value. The parser ignores connectors.

Short scale

The American numbering system: million = 10^6, billion = 10^9, trillion = 10^12. The default for this parser.

Tokenization

The first step of parsing: splitting input text into individual words for evaluation.

Decimal point keyword

The word 'point' (sometimes 'dot' in informal speech) used to indicate the decimal separator in spoken numbers.