diff --git a/docs/usage/security.md b/docs/usage/security.md
index 3c7f1e3f9ba8..5d4d429c6fed 100644
--- a/docs/usage/security.md
+++ b/docs/usage/security.md
@@ -170,7 +170,7 @@ When `--api-key` is configured, the following endpoints require Bearer token aut
 - `/v1/messages/render` - Render Anthropic-compatible messages (available on `vllm serve` only when `--enable-scale-out` is set, or on `vllm launch render`)
 - `/v1/messages/count_tokens` - Count tokens for Anthropic messages
 - `/v1/responses` - Create a response
-- `/v1/responses/render` - Render a self-contained response request (available on `vllm serve` only when `VLLM_ENABLE_SCALE_OUT_ENDPOINTS=1`, or on `vllm launch render` unless explicitly disabled)
+- `/v1/responses/render` - Render a self-contained response request (available on `vllm serve` only when `--enable-scale-out` is set, or on `vllm launch render` unless explicitly disabled)
 - `/v1/responses/{response_id}` - Retrieve a response
 - `/v1/responses/{response_id}/cancel` - Cancel a response
 - `/v1/score` - Scoring API
diff --git a/tests/entrypoints/scale_out/ec_integration/run_scale_out_ec_e2e_test.sh b/tests/entrypoints/scale_out/ec_integration/run_scale_out_ec_e2e_test.sh
index b99b1c3ca5b0..462ada22d712 100755
--- a/tests/entrypoints/scale_out/ec_integration/run_scale_out_ec_e2e_test.sh
+++ b/tests/entrypoints/scale_out/ec_integration/run_scale_out_ec_e2e_test.sh
@@ -181,8 +181,9 @@ run_scale_out_ec() {
     # Encode-only EC producer instance. It runs the vision encoder and
     # publishes embeddings through the ECExampleConnector shared storage.
     echo "Starting encode instance on GPU $GPU_E, port $ENCODE_PORT"
-    env CUDA_VISIBLE_DEVICES="$GPU_E" VLLM_ENABLE_SCALE_OUT_ENDPOINTS=1 \
+    env CUDA_VISIBLE_DEVICES="$GPU_E" \
         vllm serve "$MODEL" \
+        --enable-scale-out \
         --port "$ENCODE_PORT" \
         --max-model-len "$MAX_MODEL_LEN" \
         --enforce-eager \
@@ -204,8 +205,9 @@ run_scale_out_ec() {
     # features plus ec_transfer_params and loads the embeddings that the
     # encode instance published.
     echo "Starting prefill instance on GPU $GPU_PD, port $PREFILL_PORT"
-    env CUDA_VISIBLE_DEVICES="$GPU_PD" VLLM_ENABLE_SCALE_OUT_ENDPOINTS=1 \
+    env CUDA_VISIBLE_DEVICES="$GPU_PD" \
         vllm serve "$MODEL" \
+        --enable-scale-out \
         --port "$PREFILL_PORT" \
         --max-model-len "$MAX_MODEL_LEN" \
         --enforce-eager \