diff --git a/memori/__init__.py b/memori/__init__.py
index 59d95e88..1611fa4d 100644
--- a/memori/__init__.py
+++ b/memori/__init__.py
@@ -123,5 +123,16 @@ def set_session(self, id):
         self.config.session_id = id
         return self
 
-    def recall(self, query: str, limit: int = 5):
-        return Recall(self.config).search_facts(query, limit)
+    def recall(self, query: str, limit: int = 5, page: int = 1, per_page: int | None = None):
+        """Search for relevant facts based on query.
+        
+        Args:
+            query: Search query string
+            limit: Maximum number of results to return
+            page: Page number for pagination (1-based)
+            per_page: Number of results per page (None to disable pagination)
+        
+        Returns:
+            List of fact dictionaries with id, content, and similarity score
+        """
+        return Recall(self.config).search_facts(query, limit, page=page, per_page=per_page)
diff --git a/memori/_config.py b/memori/_config.py
index ea8dcfcf..bd15cdb6 100644
--- a/memori/_config.py
+++ b/memori/_config.py
@@ -41,6 +41,8 @@ def __init__(self):
         self.recall_embeddings_limit = 1000
         self.recall_facts_limit = 5
         self.recall_relevance_threshold = 0.1
+        self.recall_default_page = 1  # Default page number (1-based)
+        self.recall_default_per_page = 10  # Default results per page
         self.request_backoff_factor = 1
         self.request_num_backoff = 5
         self.request_secs_timeout = 5
diff --git a/memori/_search.py b/memori/_search.py
index 3539ffb2..95df690d 100644
--- a/memori/_search.py
+++ b/memori/_search.py
@@ -40,6 +40,8 @@ def find_similar_embeddings(
     embeddings: list[tuple[int, Any]],
     query_embedding: list[float],
     limit: int = 5,
+    page: int = 1,
+    per_page: int | None = None,
 ) -> list[tuple[int, float]]:
     """Find most similar embeddings using FAISS cosine similarity.
 
@@ -89,6 +91,11 @@ def find_similar_embeddings(
         if embedding_idx >= 0 and embedding_idx < len(id_list):
             results.append((id_list[embedding_idx], float(similarities[0][result_idx])))
 
+    # Apply pagination if per_page is specified
+    if per_page is not None:
+        offset = page * per_page
+        results = results[offset:offset + per_page]
+
     return results
 
 
@@ -98,6 +105,8 @@ def search_entity_facts(
     query_embedding: list[float],
     limit: int,
     embeddings_limit: int,
+    page: int = 1,
+    per_page: int | None = None,
 ) -> list[dict]:
     """Search entity facts by embedding similarity.
 
@@ -117,7 +126,7 @@ def search_entity_facts(
         return []
 
     embeddings = [(row["id"], row["content_embedding"]) for row in results]
-    similar = find_similar_embeddings(embeddings, query_embedding, limit)
+    similar = find_similar_embeddings(embeddings, query_embedding, limit, per_page=per_page)
 
     if not similar:
         return []
diff --git a/memori/memory/recall.py b/memori/memory/recall.py
index c171fb9c..282682e6 100644
--- a/memori/memory/recall.py
+++ b/memori/memory/recall.py
@@ -25,7 +25,8 @@ def __init__(self, config: Config) -> None:
         self.config = config
 
     def search_facts(
-        self, query: str, limit: int | None = None, entity_id: int | None = None
+        self, query: str, limit: int | None = None, entity_id: int | None = None,
+        page: int | None = None, per_page: int | None = None
     ) -> list[dict]:
         if self.config.storage is None or self.config.storage.driver is None:
             return []
@@ -41,6 +42,12 @@ def search_facts(
         if limit is None:
             limit = self.config.recall_facts_limit
 
+        if page is None:
+            page = self.config.recall_default_page
+        
+        if per_page is None:
+            per_page = self.config.recall_default_per_page
+
         query_embedding = embed_texts(query)[0]
 
         facts = []
@@ -52,6 +59,7 @@ def search_facts(
                     query_embedding,
                     limit,
                     self.config.recall_embeddings_limit,
+                    page=page,
                 )
                 break
             except OperationalError as e: