diff --git a/evaluation/qwen_image_inputs/768x1024.png b/evaluation/qwen_image_inputs/768x1024.png
new file mode 100644
index 000000000000..3a9ecb1eaad4
Binary files /dev/null and b/evaluation/qwen_image_inputs/768x1024.png differ
diff --git a/evaluation/run_qwen_image_edit_lora.py b/evaluation/run_qwen_image_edit_lora.py
new file mode 100644
index 000000000000..ea0b89039d06
--- /dev/null
+++ b/evaluation/run_qwen_image_edit_lora.py
@@ -0,0 +1,52 @@
+from sglang.multimodal_gen import DiffGenerator
+
+def main():
+    # Create a diff generator from a pre-trained model
+    generator = DiffGenerator.from_pretrained(
+        model_path="Qwen/Qwen-Image-Edit",
+        lora_path="lightx2v/Qwen-Image-Edit-2511-Lightning/Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors",
+        num_gpus=2,
+        ulysses_degree=2,
+        tp_size=1,
+        enable_torch_compile=True,
+        dit_cpu_offload=False,
+        text_encoder_cpu_offload=False,
+        vae_cpu_offload=False,
+        image_encoder_cpu_offload=False,
+        image_encoder_precision="bf16",
+        vae_precision="bf16",
+    )
+
+    # warm-up
+    generator.generate(
+            sampling_params_kwargs=dict(
+                prompt="make the clothes to red",
+                image_path="./qwen_image_inputs/768x1024.png",
+                output_path="./qwen_image_edit/",  # Controls where images are saved
+                height=1024,
+                width=768,
+                num_inference_steps=8,
+                seed=42,
+                guidance_scale=1, 
+                # profile=True, # profiling flag
+                # profile_all_stages=True, # profiling flag
+            )
+        )
+    # Generate the image
+    generator.generate(
+            sampling_params_kwargs=dict(
+                prompt="make the clothes to red",
+                image_path="./qwen_image_inputs/768x1024.png",
+                output_path="./qwen_image_edit/",  # Controls where images are saved
+                height=1024,
+                width=768,
+                num_inference_steps=8,
+                seed=42,
+                guidance_scale=1, 
+                # profile=True, # profiling flag
+                # profile_all_stages=True, # profiling flag
+            )
+        )
+
+if __name__ == '__main__':
+    main()
\ No newline at end of file