diff --git a/lite/api/benchmark.cc b/lite/api/benchmark.cc
index c137324b576f9f9399669a5e68d948b9921e4866..8eb76d91db07e566848fa5f7ea8d79483fc6cd3f 100644
--- a/lite/api/benchmark.cc
+++ b/lite/api/benchmark.cc
@@ -46,7 +46,6 @@ void OutputOptModel(const std::string& load_model_dir,
   config.set_model_dir(load_model_dir);
   std::vector<Place> vaild_places = {
       Place{TARGET(kARM), PRECISION(kFloat)},
-      Place{TARGET(kX86), PRECISION(kFloat)},
   };
   if (FLAGS_is_quantized_model) {
     vaild_places.insert(vaild_places.begin(),
diff --git a/lite/api/model_test.cc b/lite/api/model_test.cc
index cf5fa4981a173ceb77e091ea9be0e510eb53980a..dc9fac96ee848d73ca14c8dc4555c0f44951400a 100644
--- a/lite/api/model_test.cc
+++ b/lite/api/model_test.cc
@@ -47,7 +47,6 @@ void OutputOptModel(const std::string& load_model_dir,
   lite_api::CxxConfig config;
   config.set_model_dir(load_model_dir);
   config.set_valid_places({
-      Place{TARGET(kX86), PRECISION(kFloat)},
       Place{TARGET(kARM), PRECISION(kFloat)},
   });
   auto predictor = lite_api::CreatePaddlePredictor(config);