diff --git a/ggml/src/ggml-cuda/ggml-cuda.cu b/ggml/src/ggml-cuda/ggml-cuda.cu
index 1861f19688..de5a0bbd0f 100644
--- a/ggml/src/ggml-cuda/ggml-cuda.cu
+++ b/ggml/src/ggml-cuda/ggml-cuda.cu
@@ -2307,7 +2307,9 @@ static void ggml_cuda_mul_mat_id(ggml_backend_cuda_context & ctx, ggml_tensor *
     }
 
     // note: this path should not be reached when recording CUDA graphs, because it requires stream synchronization
-
+    cudaStreamCaptureStatus capture_status;
+    CUDA_CHECK(cudaStreamIsCapturing(stream, &capture_status));
+    GGML_ASSERT(capture_status == cudaStreamCaptureStatusNone);
     cudaStream_t stream = ctx.stream();
 
     GGML_ASSERT(nb12 % nb11 == 0);