diff --git a/ggml.c b/ggml.c
index 41557ab67..b1c76e678 100644
--- a/ggml.c
+++ b/ggml.c
@@ -6321,7 +6321,7 @@ struct ggml_tensor * ggml_flash_attn_ext(
 
     // permute(0, 2, 1, 3)
     int64_t ne[4] = { q->ne[0], q->ne[2], q->ne[1], q->ne[3] };
-    struct ggml_tensor * result = ggml_new_tensor(ctx, GGML_TYPE_F32, GGML_MAX_DIMS, ne);
+    struct ggml_tensor * result = ggml_new_tensor(ctx, GGML_TYPE_F32, 4, ne);
 
     float params[] = { scale };
     ggml_set_op_params(result, params, sizeof(params));