diff --git a/README.md b/README.md
index 15b13243b4e2f1897cfb31fd42013a6fc5ecef25..f7ca4bb2595a5d575976d9a1738c3476c7cf801d 100644
--- a/README.md
+++ b/README.md
@@ -13,7 +13,7 @@ We are able to both train and query a DNN, load it from disk or save
 it to disk.
 
 Being this machine learning application very heavy as time complexity,
-we beleave it is a good benchmark for the EPI processor.
+we believe it is a good benchmark for the EPI processor.
 
 In the future we might add other **machine learning** algorithms
-(both C++ source code, datasets and documentation).
\ No newline at end of file
+(both C++ source code, datasets and documentation).