diff --git a/pipelines/perf-eval/apiserver-benchmark-virtualnodes10-pods100.yml b/pipelines/perf-eval/apiserver-benchmark-virtualnodes10-pods100.yml
index e445033bd..e00552e7d 100644
--- a/pipelines/perf-eval/apiserver-benchmark-virtualnodes10-pods100.yml
+++ b/pipelines/perf-eval/apiserver-benchmark-virtualnodes10-pods100.yml
@@ -33,7 +33,7 @@ stages:
               extra_benchmark_subcmd_args: ""
               disable_warmup: "true"
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node10_job1_pod100
             benchmark_subcmd_args: "--total 1000"
           max_parallel: 2
@@ -58,7 +58,7 @@ stages:
               flowcontrol: "exempt:5"
               extra_benchmark_subcmd_args: ""
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node10_job1_pod100
             benchmark_subcmd_args: "--total 1000"
           max_parallel: 2
diff --git a/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods10k.yml b/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods10k.yml
index 62b503bb9..04720c022 100644
--- a/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods10k.yml
+++ b/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods10k.yml
@@ -37,7 +37,7 @@ stages:
               flowcontrol: "exempt:5"
               extra_benchmark_subcmd_args: "--padding-bytes=16384"
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node100_pod10k
             benchmark_subcmd_args: "--total 72000 --deployments=10 --interval 24h --cpu 64 --memory 192"
           max_parallel: 2
@@ -68,7 +68,7 @@ stages:
               flowcontrol: "exempt:5"
               extra_benchmark_subcmd_args: "--padding-bytes=16384"
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node100_pod10k
             benchmark_subcmd_args: "--total 72000 --deployments=10 --interval 24h --cpu 64 --memory 192"
           max_parallel: 2
diff --git a/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods3k.yml b/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods3k.yml
index 91a414ffd..2240a79b4 100644
--- a/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods3k.yml
+++ b/pipelines/perf-eval/apiserver-benchmark-virtualnodes100-pods3k.yml
@@ -31,7 +31,7 @@ stages:
               flowcontrol: "exempt:5"
               extra_benchmark_subcmd_args: ""
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node100_job1_pod3k
             benchmark_subcmd_args: "--total 36000"
           max_parallel: 2
@@ -56,7 +56,7 @@ stages:
               flowcontrol: "exempt:5"
               extra_benchmark_subcmd_args: ""
           engine_input:
-            runner_image: telescope.azurecr.io/oss/kperf:v0.1.5
+            runner_image: ghcr.io/azure/kperf:0.1.5
             benchmark_subcmd: node100_job1_pod3k
             benchmark_subcmd_args: "--total 36000"
           max_parallel: 2