diff --git a/server/text_generation_server/models/__init__.py b/server/text_generation_server/models/__init__.py
index 9613218d..026ab374 100644
--- a/server/text_generation_server/models/__init__.py
+++ b/server/text_generation_server/models/__init__.py
@@ -99,10 +99,21 @@ def get_model(
         else:
             return Galactica(model_id, revision, quantize=quantize)
 
+    if model_id.startswith("bigcode/"):
+        if sharded:
+            if not FLASH_ATTENTION:
+                raise NotImplementedError(
+                    FLASH_ATT_ERROR_MESSAGE.format(f"Sharded Santacoder")
+                )
+            return FlashSantacoderSharded(model_id, revision, quantize=quantize)
+        else:
+            santacoder_cls = FlashSantacoder if FLASH_ATTENTION else SantaCoder
+            return santacoder_cls(model_id, revision, quantize=quantize)
+
     config = AutoConfig.from_pretrained(model_id, revision=revision)
     model_type = config.model_type
 
-    if model_type == "gpt_bigcode" or model_id.startswith("bigcode/"): 
+    if model_type == "gpt_bigcode" or model_id.startswith("bigcode/"):
         if sharded:
             if not FLASH_ATTENTION:
                 raise NotImplementedError(