diff --git a/docs/source/en/processors.md b/docs/source/en/processors.md
index 7bd6be74f5b..bb2a149ce62 100644
--- a/docs/source/en/processors.md
+++ b/docs/source/en/processors.md
@@ -106,6 +106,8 @@ dataset[0]["text"]
 Remember to resample the sampling rate to match the pretrained models required sampling rate.
 
 ```py
+from datasets import Audio
+
 dataset = dataset.cast_column("audio", Audio(sampling_rate=16000))
 ```