diff --git a/.github/workflows/unit_tests.yml b/.github/workflows/unit_tests.yml
index 5e22da68c..463f678e3 100644
--- a/.github/workflows/unit_tests.yml
+++ b/.github/workflows/unit_tests.yml
@@ -562,6 +562,11 @@ jobs:
           runner_q=$(urlencode "$RUNNER")
           exclusive_q=$(urlencode "$EXCLUSIVE")
           
+          count=1
+          if [[ "${{ matrix.test_script }}" == *test_multi_gpu_inference* ]]; then
+            count=5
+          fi
+          
           start_s=$(date +%s)
           gpu_id=""
           
@@ -571,7 +576,7 @@ jobs:
           while true; do
             # Refresh timestamp every request to avoid cache or duplicated requests
             ts_ms=$(date +%s%3N)
-            url="$BASE_URL/gpu/get?runid=$RUN_ID&timestamp=$ts_ms&test=$test_q&runner=$runner_q&exclusive=$exclusive_q"
+            url="$BASE_URL/gpu/get?runid=$RUN_ID&timestamp=$ts_ms&test=$test_q&runner=$runner_q&exclusive=$exclusive_q&count=$count"
             echo "requesting GPU with: $url"
           
             # Call allocator