diff --git a/jarvis/config/spacy.json b/jarvis/config/spacy.json
new file mode 100644
index 0000000..44fb619
--- /dev/null
+++ b/jarvis/config/spacy.json
@@ -0,0 +1,19 @@
+{
+  "fr-fr": "fr_core_news_sm",
+  "en-en": "en_core_web_sm",
+  "en-us": "en_core_web_sm",
+  "de-de": "de_core_news_sm",
+  "tr-tr": "xx_ent_wiki_sm",
+  "cs-cz": "xx_ent_wiki_sm",
+  "el-gr": "xx_ent_wiki_sm",
+  "da-dk": "da_core_news_sm",
+  "et-ee": "xx_ent_wiki_sm",
+  "pt-pt": "pt_core_news_sm",
+  "es-es": "es_core_news_sm",
+  "nl-nl": "nb_core_news_sm",
+  "fi-fi": "xx_ent_wiki_sm",
+  "it-it": "it_core_news_sm",
+  "pl-pl": "pl_core_news_sm",
+  "sl-si": "xx_ent_wiki_sm",
+  "sv-se": "xx_ent_wiki_sm"
+}
\ No newline at end of file
diff --git a/jarvis/utils/languages_utils.py b/jarvis/utils/languages_utils.py
index 01f1694..47fd563 100644
--- a/jarvis/utils/languages_utils.py
+++ b/jarvis/utils/languages_utils.py
@@ -26,3 +26,15 @@ def get_language_full_name(name=None):
         return config_json.get(name)
 
     return 'english'
+
+
+def get_spacy_model(language=None):
+    spacy_model = json.load(open(path + "/config/spacy.json", encoding='utf-8', mode='r'))
+
+    if language is None:
+        language = get_language()
+
+    if language in spacy_model:
+        return spacy_model.get(language)
+
+    return 'xx_ent_wiki_sm'  # multi-language model (for unsupported languages)
diff --git a/jarvis/utils/nlp_utils.py b/jarvis/utils/nlp_utils.py
index ec79e77..b4994be 100644
--- a/jarvis/utils/nlp_utils.py
+++ b/jarvis/utils/nlp_utils.py
@@ -1,13 +1,10 @@
 import spacy
-from nltk.corpus import stopwords
+
+from jarvis.utils import languages_utils
 
 
 def get_spacy_nlp():
-    """
-
-    :return: spacy
-    """
-    nlp = spacy.load("en_core_web_sm")
+    nlp = spacy.load(languages_utils.get_spacy_model())
     return nlp
 
 
@@ -18,10 +15,3 @@ def get_text_without_stopwords(sentence):
     filtered_sentence = [w for w in sentence.lower().split() if w not in stop_words]
     filtered_sentence = " ".join(filtered_sentence)
     return filtered_sentence
-
-
-def get_text_without_stopwords_nltk(sentence, language='english'):
-    stop_words = set(stopwords.words(language))
-    filtered_sentence = [w for w in sentence.lower().split() if w not in stop_words]
-    filtered_sentence = " ".join(filtered_sentence)
-    return filtered_sentence
diff --git a/requirements.txt b/requirements.txt
index ea8d29a..c61ee7f 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -4,4 +4,5 @@ nltk~=3.6.2
 torch~=1.9.0
 numpy~=1.21.1
 requests~=2.26.0
-adapt-parser
\ No newline at end of file
+adapt-parser
+spacy~=3.1.1
\ No newline at end of file