diff --git a/meilisearch-core/src/serde/extract_document_id.rs b/meilisearch-core/src/serde/extract_document_id.rs
index 5c2a81d2e..01427c5da 100644
--- a/meilisearch-core/src/serde/extract_document_id.rs
+++ b/meilisearch-core/src/serde/extract_document_id.rs
@@ -23,7 +23,13 @@ pub fn value_to_string(value: &Value) -> Option<String> {
         Value::Null => None,
         Value::Bool(_) => None,
         Value::Number(value) => Some(value.to_string()),
-        Value::String(value) => Some(value.to_string()),
+        Value::String(value) => {
+            if value.chars().all(|x| x.is_ascii_alphanumeric() || x == '-' || x == '_') {
+                Some(value.to_string())
+            } else {
+                None
+            }
+        },
         Value::Array(_) => None,
         Value::Object(_) => None,
     }
diff --git a/meilisearch-core/src/serde/mod.rs b/meilisearch-core/src/serde/mod.rs
index c25c4b158..0469f352d 100644
--- a/meilisearch-core/src/serde/mod.rs
+++ b/meilisearch-core/src/serde/mod.rs
@@ -55,7 +55,7 @@ impl fmt::Display for SerializerError {
                 f.write_str("serialized document does not have an id according to the schema")
             }
             SerializerError::InvalidDocumentIdType => {
-                f.write_str("document identifier can only be of type string or number")
+                f.write_str("document identifier can only be of type number or string (A-Z, a-z, 0-9, -_)")
             }
             SerializerError::Zlmdb(e) => write!(f, "heed related error: {}", e),
             SerializerError::SerdeJson(e) => write!(f, "serde json error: {}", e),
diff --git a/meilisearch-http/src/routes/document.rs b/meilisearch-http/src/routes/document.rs
index 7c91219b8..ebb3851be 100644
--- a/meilisearch-http/src/routes/document.rs
+++ b/meilisearch-http/src/routes/document.rs
@@ -117,13 +117,14 @@ pub async fn get_all_documents(ctx: Context<Data>) -> SResult<Response> {
     Ok(tide::response::json(response_body))
 }
 
-fn infered_schema(document: &IndexMap<String, Value>) -> Option<meilisearch_schema::Schema> {
+fn infered_schema(document: &IndexMap<String, Value>, identifier: Option<String>) -> Option<meilisearch_schema::Schema> {
     use meilisearch_schema::{SchemaBuilder, DISPLAYED, INDEXED};
 
-    let mut identifier = None;
+    let mut identifier = identifier;
     for key in document.keys() {
         if identifier.is_none() && key.to_lowercase().contains("id") {
-            identifier = Some(key);
+            identifier = Some(key.to_string());
+            break;
         }
     }
 
@@ -139,12 +140,21 @@ fn infered_schema(document: &IndexMap<String, Value>) -> Option<meilisearch_sche
     }
 }
 
+#[derive(Default, Deserialize)]
+#[serde(deny_unknown_fields)]
+struct UpdateDocumentsQuery {
+    identifier: Option<String>,
+}
+
 async fn update_multiple_documents(mut ctx: Context<Data>, is_partial: bool) -> SResult<Response> {
     ctx.is_allowed(DocumentsWrite)?;
 
+    let index = ctx.index()?;
+
     let data: Vec<IndexMap<String, Value>> =
         ctx.body_json().await.map_err(ResponseError::bad_request)?;
-    let index = ctx.index()?;
+    let query: UpdateDocumentsQuery = ctx
+        .url_query().unwrap_or_default();
 
     let db = &ctx.state().db;
     let reader = db.main_read_txn().map_err(ResponseError::internal)?;
@@ -155,7 +165,7 @@ async fn update_multiple_documents(mut ctx: Context<Data>, is_partial: bool) ->
         .schema(&reader)
         .map_err(ResponseError::internal)?;
     if current_schema.is_none() {
-        match data.first().and_then(infered_schema) {
+        match data.first().and_then(|docs| infered_schema(docs, query.identifier)) {
             Some(schema) => {
                 index
                     .schema_update(&mut update_writer, schema)
diff --git a/meilisearch-http/src/routes/mod.rs b/meilisearch-http/src/routes/mod.rs
index ee6d75e47..e27d2887b 100644
--- a/meilisearch-http/src/routes/mod.rs
+++ b/meilisearch-http/src/routes/mod.rs
@@ -72,7 +72,7 @@ pub fn load_routes(app: &mut tide::App<Data>) {
                     });
 
                     router
-                        .at("/delete")
+                        .at("/delete-batch")
                         .post(document::delete_multiple_documents);
                 });