A blazing-fast, zero-dependency, arbitrary-precision decimal math library for TypeScript and JavaScript. Built on native BigInt to safely handle massive numbers, exact decimals, and eliminate floating-point errors.
Perfect for financial applications, trading platforms, and anywhere precise calculation is critical.
npm install bigint-decimal
# or
yarn add bigint-decimal
# or
pnpm add bigint-decimal
Native JavaScript floating-point math is notoriously imprecise:
console.log(0.1 + 0.2); // 0.30000000000000004 ❌bigint-decimal fixes this natively, without the bloat of massive external math libraries:
import { BigintDecimal } from 'bigint-decimal';
const a = new BigintDecimal("0.1");
const b = new BigintDecimal("0.2");
console.log(a.plus(b)); // "0.3" ✅BigintDecimal supports all standard arithmetic operations. Always pass numbers as strings to guarantee precision before JavaScript evaluates them.
import { BigintDecimal } from 'bigint-decimal';
const walletBalance = new BigintDecimal("1500.50");
const tradeAmount = new BigintDecimal("200.75");
const gasFee = new BigintDecimal("0.005");
// Addition
const newBalance = walletBalance.plus(tradeAmount); // "1701.25"
// Subtraction
const finalBalance = newBalance.minus(gasFee); // "1701.245"
// Multiplication
const price = new BigintDecimal("45000.50");
const amount = new BigintDecimal("0.5");
const totalValue = price.times(amount); // "22500.25"
// Division
const totalTokens = new BigintDecimal("1000");
const splits = new BigintDecimal("3");
const perPerson = totalTokens.div(splits); // "333.33333333333333333333"Every math operation returns a new BigintDecimal instance, allowing you to cleanly chain complex formulas.
const result = new BigintDecimal("100")
.plus("50.5")
.times("2")
.minus("1.1")
.div("2");
console.log(result); // "149.95"Easily compare values without precision loss.
const requiredMargin = new BigintDecimal("1000.00");
const userBalance = new BigintDecimal("999.99");
userBalance.gt(requiredMargin); // false (Greater Than)
userBalance.gte(requiredMargin); // false (Greater Than or Equal)
userBalance.lt(requiredMargin); // true (Less Than)
userBalance.lte(requiredMargin); // true (Less Than or Equal)
userBalance.eq("999.99"); // true (Equal)Round to a specific number of decimal places. Standard rounding rules apply (halves round up).
const fee = new BigintDecimal("1.555");
console.log(fee.round(2)); // "1.56"
console.log(fee.round(0)); // "2"Division operations can result in infinitely repeating decimals. BigintDecimal handles this by defaulting to a maximum precision of 20 decimal places. You can adjust this globally or per instance.
// Change instance precision
const exactDiv = new BigintDecimal("1");
exactDiv.setPrecision(5);
console.log(exactDiv.div("3")); // "0.33333"
// Change global precision for all new instances
BigintDecimal.setPrecisionGlobal(2);
console.log(new BigintDecimal("1").div("3")); // "0.33"BigintDecimal safely parses a wide variety of inputs, including scientific notation and massive integers.
new BigintDecimal("1e-18"); // "0.000000000000000001"
new BigintDecimal("9999999999999999999999999999"); // Handles massive numbers natively
new BigintDecimal("-0.0005"); // Negative floats
new BigintDecimal(100n); // Native BigIntNote: While passing number primitives (like new BigintDecimal(100)) is supported, passing them as strings ("100") is strongly recommended to avoid JS evaluating floating points before the class receives them.
MIT