diff --git a/diseases-ner/README.md b/diseases-ner/README.md
new file mode 100644
index 0000000..6607594
--- /dev/null
+++ b/diseases-ner/README.md
@@ -0,0 +1,35 @@
+# chem-ner
+
+Cette instance propose un outil de reconnaissance d'entités nommées de maladies.
+
+## Configuration
+
+L'application à utiliser est XXX.
+
+## Utilisation
+
+- [v1/diseases/tagger](#v1)
+
+### v1
+
+Ce web-service renvoie la liste des entités nommées en chimie présentes dans le texte.
+
+Il prend en entrée du JSON avec deux champs, `id` et `value`, et renvoie un JSON avec le la liste des entités en fonction de leur étiquettes dans le champ `value`.
+
+#### Exemple de v1/first-name/gender
+
+Entrée
+
+```bash
+$ cat <<EOF | curl -X POST --data-binary @- "?????????"
+[{"id": "1", "value":"They have been used in the prevention and treatment of malaria and autoimmune diseases, including systemic lupus erythematosus and rheumatoid arthritis."}]
+EOF
+```
+
+Sortie
+
+```json
+[
+  {"id": 1, "value": {"diseases": ["malaria", "autoimmune diseases", "systemic lupus erythematosus", "rheumatoid arthritis"]}}
+]
+```
diff --git a/diseases-ner/examples.http b/diseases-ner/examples.http
new file mode 100644
index 0000000..1d059f6
--- /dev/null
+++ b/diseases-ner/examples.http
@@ -0,0 +1,18 @@
+# Ces exemples peuvent être exécutés directement dans VSCode, en utilisant l'extension REST Client (humao.rest-client)
+# Décommenter/commenter les lignes voulues pour tester localement
+# @baseUrl=http://localhost:31976
+@baseUrl=http://diseases-ner.tdmservices.intra.inist.fr/
+# @baseUrl=https://diseases-ner.services.istex.fr/
+
+###
+# @name v1AstroNer
+# Reconnaissance d'entités nommées de chimie
+POST {{baseUrl}}/v1/diseases/tagger?indent=true HTTP/1.1
+Content-Type: application/json
+
+[
+  {
+    "id": 1,
+    "value":"They have been used in the prevention and treatment of malaria and autoimmune diseases, including systemic lupus erythematosus and rheumatoid arthritis."
+  }
+]
diff --git a/diseases-ner/swagger.json b/diseases-ner/swagger.json
new file mode 100644
index 0000000..be733f0
--- /dev/null
+++ b/diseases-ner/swagger.json
@@ -0,0 +1,33 @@
+{
+    "openapi": "3.0.0",
+    "info": {
+      "title": "diseases-ner - Reconnaissance d'entités nommées de maladies",
+      "summary": "Reconnaissance des entités nommées de maladies",
+      "version": "1.0.6",
+      "termsOfService": "https://services.istex.fr/",
+      "contact": {
+        "name": "Inist-CNRS",
+        "url": "https://www.inist.fr/nous-contacter/"
+      }
+    },
+    "servers": [
+      {
+        "x-comment": "Will be automatically completed by the ezs server."
+      },
+      {
+        "url": "http://vptdmservices.intra.inist.fr:49240/",
+        "description": "Latest version for production",
+        "x-profil": "Standard"
+      }
+    ],
+    "tags": [
+      {
+        "name": "chem-ner",
+        "description": "Reconnaissance d'entités nommées en chimie",
+        "externalDocs": {
+          "description": "Plus de documentation",
+          "url": "https://gitbucket.inist.fr/tdm/web-services/chem-ner/README.md"
+        }
+      }
+    ]
+  }
\ No newline at end of file
diff --git a/diseases-ner/v1/diseases/models.dvc b/diseases-ner/v1/diseases/models.dvc
new file mode 100644
index 0000000..b6c0e73
--- /dev/null
+++ b/diseases-ner/v1/diseases/models.dvc
@@ -0,0 +1,6 @@
+outs:
+- md5: ddf4fda3fe13cd7c131205d4e34b6de9.dir
+  size: 432104605
+  nfiles: 11
+  hash: md5
+  path: models
diff --git a/diseases-ner/v1/diseases/tagger.ini b/diseases-ner/v1/diseases/tagger.ini
new file mode 100644
index 0000000..cd5deb4
--- /dev/null
+++ b/diseases-ner/v1/diseases/tagger.ini
@@ -0,0 +1,42 @@
+# OpenAPI Documentation - JSON format (dot notation)
+mimeType = application/json
+
+post.description = Trouve des entités nommées de maladies dans un texte
+post.responses.default.description = Renvoie un Json composé comportant un champs `value` correspondant aux entités de maladies trouvées
+post.responses.default.content.application/json.schema.$ref =  #/components/schemas/JSONStream
+post.summary = Recherche d'entité nommées de maladies
+post.requestBody.required = true
+post.requestBody.content.application/json.schema.$ref = #/components/schemas/JSONStream
+post.parameters.0.in = query
+post.parameters.0.name = indent
+post.parameters.0.schema.type = boolean
+post.parameters.0.description = Indent or not the JSON Result
+
+# Examples
+
+post.requestBody.content.application/json.example.0.id: 1
+post.requestBody.content.application/json.example.0.value: They have been used in the prevention and treatment of malaria and autoimmune diseases, including systemic lupus erythematosus and rheumatoid arthritis.
+post.responses.default.content.application/json.example.0.id: 1
+post.responses.default.content.application/json.example.0.value:
+post.responses.default.content.application/json.example.0.value.diseases: ["malaria", "autoimmune diseases", "systemic lupus erythematosus", "rheumatoid arthritis"]
+
+
+[use]
+# exec
+plugin = @ezs/spawn
+# JSONParse
+plugin = @ezs/basics
+
+[JSONParse]
+separator = *
+
+[expand]
+path = value
+size = 100
+
+[expand/exec]
+# command should be executable !
+command = ./v1/diseases/tagger.py
+
+[dump]
+indent = env('indent', false)
diff --git a/diseases-ner/v1/diseases/tagger.py b/diseases-ner/v1/diseases/tagger.py
new file mode 100644
index 0000000..beaa50f
--- /dev/null
+++ b/diseases-ner/v1/diseases/tagger.py
@@ -0,0 +1,109 @@
+#!/opt/bitnami/python/bin/python3.7
+# -*- coding: utf-8 -*-
+
+import torch
+from transformers import AutoTokenizer, AutoModelForTokenClassification
+import sys
+import json
+import re
+import unicodedata
+
+
+# Normalize 
+# Normalisation du texte :
+def remove_accents(text):
+    text = unicodedata.normalize("NFD", text)
+    text = re.sub("[\u0300-\u036f]", "", text)
+    return text
+
+def normalizeText(text):
+    text = text.lower()
+    text = remove_accents(text).replace(" ","")
+    return text
+
+
+## Predicts developped formulas
+# Load model
+tokenizer = AutoTokenizer.from_pretrained('./v1/diseases/models')
+model = AutoModelForTokenClassification.from_pretrained('./v1/diseases/models', config='./v1/diseases/models/config.json')
+
+
+# predicts text
+def predict_formula_ml(input_text):
+    #tokenizer 
+    tokens = tokenizer(input_text, return_tensors='pt', truncation=True, max_length=512)
+
+    # Predicts
+    with torch.no_grad():
+        output = model(**tokens)
+
+    predictions = torch.argmax(output.logits, dim=-1)
+
+    # Get token that contains "disease"
+    tokens = tokenizer.convert_ids_to_tokens(tokens['input_ids'][0])
+    disease_tokens_list = []
+    i=0
+
+    while i < len(predictions[0]):
+        # prediction [0][i] depends of i : {0 : "B-disease" , 1 : "I-disease" , 2: "NOT a disease NE"}
+        k=0
+        if predictions[0][i] < 2:
+            disease_tokens_toappend = []
+            while predictions[0][i+k] < 2:
+                disease_tokens_toappend.append(tokens[i+k])
+                k+=1
+            disease_tokens_list.append(disease_tokens_toappend)
+        i+=k+1
+    value = []
+    for disease_tokens in disease_tokens_list:
+        value.append(tokenizer.decode(tokenizer.convert_tokens_to_ids(disease_tokens)))
+    return value
+
+# if text too long
+def split_text(text):
+    if len(text)>=512:
+        text_split = text.split('.')
+    else:
+        text_split = [text]
+    return text_split
+
+# predicts text after, either it is splitted or not
+def predict_formula_ml_list(list):
+    output = []
+    for elt in list:
+        output+= predict_formula_ml(elt)
+    return output
+
+# remove bad space in outputs
+def curate_list(input_list):
+    output_list = []
+    for elt in input_list:
+        if '#' not in elt:
+            output_list.append(
+                        elt.replace('- ','-').replace(' -','-').replace('( ','(').replace(' (','(').replace(') ',')').replace(' )',')').replace('[ ','[')
+                        .replace(' [','[').replace('] ',']').replace(' ]',']')
+                        )
+    return output_list
+
+#Disambigusate formulas :
+
+#preprocessing : remove duplicates elements
+def remove_duplicates(input_list):
+    output_list = []
+    normalized_list = []
+    for elt in input_list:
+        if normalizeText(elt) not in normalized_list:
+            output_list.append(elt)
+            normalized_list.append(normalizeText(elt))
+    return output_list
+
+
+# beginning of the ws  
+for line in sys.stdin:
+    data = json.loads(line)
+    # Use the model to find NER
+    value = remove_duplicates(curate_list(predict_formula_ml_list(split_text(data["value"]))))
+    # Standardization
+    data["value"] = {"diseases":value} # remove_duplicates(value)
+    json.dump(data, sys.stdout, ensure_ascii=False)
+    sys.stdout.write("\n")