diff --git a/README.md b/README.md
index 350ad98..456cc56 100644
--- a/README.md
+++ b/README.md
@@ -4,7 +4,7 @@ Objix is a delightfully convienient, high performance, zero dependency, lightwei
 
 The functions are all non enumerable and include copies of Object class methods and Array prototype methods applied to the values of the object as well others to delete keys, stringify, promisify, memoize, compare, split and join objects, check types, log messages and trapping/observing property assignments.
 
-This library is highly optimised with zero copy operations where possible. The un-minified source is only 3.6kb (2.7kb minified) which allows for fast loading and easy integration without additonal compilation or tree shaking. There is however limited type checking to guard against unwanted side effects, and there may be some undiscovered edge case that do not behave as expected. Performance in most cases is signifantly faster than `lodash` equivalents especially when working with small objects. For example `ob.map(fn)` is typically over 65% faster than `_.mapValues(ob, fn)` and some operations such as `filter` can be several thousand times quicker according to simple [benchmarks](docs/bench.md).
+This library is highly optimised with zero copy operations where possible. The un-minified source is only 3.6kb (2.7kb minified) which allows for fast loading and easy integration without additonal compilation or tree shaking. There is however limited type checking to guard against unwanted side effects, and there may be some undiscovered edge case that do not behave as expected. Performance in most cases is significantly faster than `lodash` equivalents especially when working with small objects. For example `ob.map(fn)` is typically over 65% faster than `_.mapValues(ob, fn)` and some operations such as `filter` can be several thousand times quicker according to simple [benchmarks](docs/bench.md).
 
 Interactive docs and demos are availble on https://objix.dev/#/api.