diff --git a/fastdeploy/model_executor/utils.py b/fastdeploy/model_executor/utils.py
index 8e90fb80fd..daf0c10f72 100644
--- a/fastdeploy/model_executor/utils.py
+++ b/fastdeploy/model_executor/utils.py
@@ -199,3 +199,19 @@ def temporary_dtype(dtype: str):
         yield
     finally:
         paddle.set_default_dtype(orig_dtype)
+
+
+def is_paddle_support_v1_loader():
+    src_shape = [32, 32]
+    tgt_shape = [1, 32, 64]
+    src_tensor = paddle.ones(src_shape, dtype="float32")
+    tgt_tensor = paddle.zeros(tgt_shape, dtype="float32")
+    for exp_id in range(tgt_shape[0]):
+        # gate
+        gate_tgt = tgt_tensor[exp_id][..., : tgt_shape[2] // 2]
+        gate_tgt.copy_(src_tensor, False)
+        # up
+        up_tgt = tgt_tensor[exp_id][..., tgt_shape[2] // 2 :]
+        up_tgt.copy_(src_tensor, False)
+    is_same = bool(paddle.all(tgt_tensor == 1))
+    return is_same
diff --git a/fastdeploy/worker/worker_process.py b/fastdeploy/worker/worker_process.py
index 2b2ed9a591..43384f8ba2 100644
--- a/fastdeploy/worker/worker_process.py
+++ b/fastdeploy/worker/worker_process.py
@@ -43,6 +43,7 @@
 from fastdeploy.inter_communicator import EngineWorkerQueue as TaskQueue
 from fastdeploy.inter_communicator import ExistTaskStatus, IPCSignal, ModelWeightsStatus
 from fastdeploy.model_executor.layers.quantization import get_quantization_config
+from fastdeploy.model_executor.utils import is_paddle_support_v1_loader
 from fastdeploy.platforms import current_platform
 from fastdeploy.utils import get_logger, parse_quantization
 from fastdeploy.worker.worker_base import WorkerBase
@@ -809,6 +810,9 @@ def initialize_fd_config(args, ranks: int = 1, local_rank: int = 0) -> FDConfig:
     update_fd_config_for_mm(fd_config)
     update_think_end_id_for_ernie(fd_config)
 
+    if load_config.load_choices == "default_v1" and not is_paddle_support_v1_loader():
+        raise ValueError("The install Paddle don't support v1 loader.")
+
     return fd_config