diff --git a/website/docs/r/mks_nodegroup_v1.html.markdown b/website/docs/r/mks_nodegroup_v1.html.markdown
index d2bb1755..917226d8 100644
--- a/website/docs/r/mks_nodegroup_v1.html.markdown
+++ b/website/docs/r/mks_nodegroup_v1.html.markdown
@@ -58,7 +58,7 @@ resource "selectel_mks_nodegroup_v1" "nodegroup_1" {
* `nodes_count` - (Required) Number of worker nodes in the node group. Changing this resizes the node group if `enable_autoscale` is false.
-* `install_nvidia_device_plugin` - (Required) Boolean flag indicates if nvidia device plugin and GPU drivers installation was requested. Positive value is only suitable for `flavor_id` with GPU. In other cases with positive value - 400 (Bad request) status code will be returned from API.
+* `install_nvidia_device_plugin` - (Required) Enables or disables installation of the NVIDIA Device Plugin and GPU drivers. Boolean flag: `true` — for flavors with GPU enables installation of the NVIDIA Device Plugin and GPU drivers. `false` — for flavors without GPU and flavors with GPU disables installation of the NVIDIA Device Plugin and GPU drivers. Learn more about [manual installation of GPU drivers](https://docs.selectel.ru/en/cloud/managed-kubernetes/node-groups/install-gpu-drivers/).
* `cpus` - (Optional) Number of vCPUs for each node. Can be skipped only when `flavor_id` is set. Changing this creates a new node group. Learn more about [Configurations](https://docs.selectel.ru/en/cloud/managed-kubernetes/node-groups/configurations/).