diff --git a/setup.py b/setup.py
index f5d45f2..1025e24 100644
--- a/setup.py
+++ b/setup.py
@@ -31,7 +31,6 @@
         'scikit-learn >= 1.2.0',
         'gensim >= 4.0.0',
         'umap-learn >= 0.5.1',
-        'hdbscan >= 0.8.27',
         'wordcloud',
     ],
     extras_require={
diff --git a/top2vec/Top2Vec.py b/top2vec/Top2Vec.py
index 712169c..f278796 100644
--- a/top2vec/Top2Vec.py
+++ b/top2vec/Top2Vec.py
@@ -9,11 +9,10 @@
 from gensim.parsing.preprocessing import strip_tags
 from gensim.models.phrases import Phrases
 import umap
-import hdbscan
 from wordcloud import WordCloud
 import matplotlib.pyplot as plt
 from joblib import dump, load
-from sklearn.cluster import dbscan
+from sklearn.cluster import dbscan, HDBSCAN
 import tempfile
 from sklearn.feature_extraction.text import CountVectorizer
 from sklearn.preprocessing import normalize
@@ -1384,7 +1383,7 @@ def compute_topics(self,
             labels = cluster.fit_predict(umap_embedding)
 
         else:
-            cluster = hdbscan.HDBSCAN(**hdbscan_args).fit(umap_embedding)
+            cluster = HDBSCAN(**hdbscan_args).fit(umap_embedding)
             labels = cluster.labels_
 
         # calculate topic vectors from dense areas of documents