diff --git a/pilot/configs/model_config.py b/pilot/configs/model_config.py
index 09b428327..2a9d48346 100644
--- a/pilot/configs/model_config.py
+++ b/pilot/configs/model_config.py
@@ -34,7 +34,8 @@ LLM_MODEL_CONFIG = {
     "chatglm-6b-int4": os.path.join(MODEL_PATH, "chatglm-6b-int4"),
     "chatglm-6b": os.path.join(MODEL_PATH, "chatglm-6b"),
     "text2vec-base": os.path.join(MODEL_PATH, "text2vec-base-chinese"),
-    "sentence-transforms": os.path.join(MODEL_PATH, "all-MiniLM-L6-v2"),
+    "sentence-transforms": os.path.join(MODEL_PATH, "all-MiniLM-L6-v2"0),
+    "guanaco-33b-merged": os.path.join(MODEL_PATH, "guanaco-33b-merged"
 }
 
 # Load model config
diff --git a/pilot/model/adapter.py b/pilot/model/adapter.py
index 83fad3d5f..208d9f438 100644
--- a/pilot/model/adapter.py
+++ b/pilot/model/adapter.py
@@ -82,6 +82,21 @@ class ChatGLMAdapater(BaseLLMAdaper):
             )
             return model, tokenizer
 
+        
+class GuanacoAdapter(BaseLLMAdaper):
+    """TODO Support guanaco"""
+
+    def match(self, model_path: str):
+        return "guanaco" in model_path
+
+    def loader(self, model_path: str, from_pretrained_kwargs: dict):
+        tokenizer = LlamaTokenizer.from_pretrained(model_path)
+        model = AutoModelForCausalLM.from_pretrained(
+            model_path, load_in_4bit=True, device_map={"": 0}, **from_pretrained_kwargs
+        )
+        return model, tokenizer
+
+
 
 class CodeGenAdapter(BaseLLMAdaper):
     pass
@@ -122,6 +137,7 @@ class GPT4AllAdapter(BaseLLMAdaper):
 
 register_llm_model_adapters(VicunaLLMAdapater)
 register_llm_model_adapters(ChatGLMAdapater)
+register_llm_model_adapters(GuanacoAdapter)
 # TODO Default support vicuna, other model need to tests and Evaluate
 
 register_llm_model_adapters(BaseLLMAdaper)
diff --git a/pilot/model/guanaco_llm.py b/pilot/model/guanaco_stream_llm.py
similarity index 94%
rename from pilot/model/guanaco_llm.py
rename to pilot/model/guanaco_stream_llm.py
index 627179dc5..8f72699d1 100644
--- a/pilot/model/guanaco_llm.py
+++ b/pilot/model/guanaco_stream_llm.py
@@ -3,7 +3,8 @@ from threading import Thread
 from transformers import TextIteratorStreamer, StoppingCriteriaList, StoppingCriteria
 
 
-def guanaco_generate_output(model, tokenizer, params, device, context_len=2048):
+
+def guanaco_stream_generate_output(model, tokenizer, params, device, context_len=2048):
     """Fork from: https://github.com/KohakuBlueleaf/guanaco-lora/blob/main/generate.py"""
     tokenizer.bos_token_id = 1
     print(params)
diff --git a/pilot/server/chat_adapter.py b/pilot/server/chat_adapter.py
index b7e102be3..400b27697 100644
--- a/pilot/server/chat_adapter.py
+++ b/pilot/server/chat_adapter.py
@@ -59,6 +59,16 @@ class ChatGLMChatAdapter(BaseChatAdpter):
 
         return chatglm_generate_stream
 
+class GuanacoChatAdapter(BaseChatAdpter):
+    """Model chat adapter for Guanaco"""
+
+    def match(self, model_path: str):
+        return "guanaco" in model_path
+
+    def get_generate_stream_func(self):
+        from pilot.model.llm_out.guanaco_stream_llm import guanaco_stream_generate_output
+
+        return guanaco_generate_output
 
 class CodeT5ChatAdapter(BaseChatAdpter):
 
@@ -86,5 +96,6 @@ class CodeGenChatAdapter(BaseChatAdpter):
 
 register_llm_model_chat_adapter(VicunaChatAdapter)
 register_llm_model_chat_adapter(ChatGLMChatAdapter)
+register_llm_model_chat_adapter(GuanacoChatAdapter)
 
 register_llm_model_chat_adapter(BaseChatAdpter)