diff --git a/opencompass/configs/models/gemma/vllm_gemma_3_12b_it.py b/opencompass/configs/models/gemma/vllm_gemma_3_12b_it.py
new file mode 100644
index 00000000..2914640f
--- /dev/null
+++ b/opencompass/configs/models/gemma/vllm_gemma_3_12b_it.py
@@ -0,0 +1,16 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='gemma-3-12b-it-vllm',
+        path='google/gemma-3-12b-it',
+        model_kwargs=dict(tensor_parallel_size=4,
+                          # for long context
+                          rope_scaling={'factor': 8.0, 'rope_type': 'linear'}),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]
diff --git a/opencompass/configs/models/gemma/vllm_gemma_3_27b_it.py b/opencompass/configs/models/gemma/vllm_gemma_3_27b_it.py
new file mode 100644
index 00000000..b6f4b93b
--- /dev/null
+++ b/opencompass/configs/models/gemma/vllm_gemma_3_27b_it.py
@@ -0,0 +1,16 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='gemma-3-27b-it-vllm',
+        path='google/gemma-3-27b-it',
+        model_kwargs=dict(tensor_parallel_size=4,
+                          # for long context
+                          rope_scaling={'factor': 8.0, 'rope_type': 'linear'}),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )   
+]
diff --git a/opencompass/configs/models/gemma/vllm_gemma_3_4b_it.py b/opencompass/configs/models/gemma/vllm_gemma_3_4b_it.py
new file mode 100644
index 00000000..22516ff7
--- /dev/null
+++ b/opencompass/configs/models/gemma/vllm_gemma_3_4b_it.py
@@ -0,0 +1,17 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='gemma-3-4b-it-vllm',
+        path='google/gemma-3-4b-it',
+        model_kwargs=dict(tensor_parallel_size=2, 
+                          # for long context
+                          rope_scaling={'factor': 8.0, 'rope_type': 'linear'}),
+        max_seq_len=140000,
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=2),
+    )
+]
diff --git a/opencompass/configs/models/hf_internlm/lmdeploy_internlm3_8b_instruct_128k.py b/opencompass/configs/models/hf_internlm/lmdeploy_internlm3_8b_instruct_128k.py
new file mode 100644
index 00000000..1cc4e251
--- /dev/null
+++ b/opencompass/configs/models/hf_internlm/lmdeploy_internlm3_8b_instruct_128k.py
@@ -0,0 +1,19 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='internlm3-8b-instruct-turbomind',
+        path='internlm/internlm3-8b-instruct',
+        engine_config=dict(session_len=142000, max_batch_size=1, tp=2,
+                           # for long context
+                           rope_scaling_factor=6.0),
+        gen_config=dict(
+            top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=8192
+        ),
+        max_seq_len=142000,
+        max_out_len=8192,
+        batch_size=1,
+        run_cfg=dict(num_gpus=2),
+    )
+]
diff --git a/opencompass/configs/models/hf_internlm/lmdeploy_oreal_32b.py b/opencompass/configs/models/hf_internlm/lmdeploy_oreal_32b.py
new file mode 100644
index 00000000..1d10bd94
--- /dev/null
+++ b/opencompass/configs/models/hf_internlm/lmdeploy_oreal_32b.py
@@ -0,0 +1,20 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+from opencompass.utils.text_postprocessors import extract_non_reasoning_content
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='OREAL-32B',
+        path='internlm/OREAL-32B',
+        engine_config=dict(session_len=32768, max_batch_size=16, tp=4),
+        gen_config=dict(top_k=1,
+                        temperature=1e-6,
+                        top_p=0.9,
+                        max_new_tokens=32768),
+        max_seq_len=32768,
+        max_out_len=32768,
+        batch_size=16,
+        run_cfg=dict(num_gpus=4),
+        pred_postprocessor=dict(type=extract_non_reasoning_content)
+    )
+]
diff --git a/opencompass/configs/models/qwen2_5/vllm_qwen2_5_14b_instruct_128k.py b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_14b_instruct_128k.py
new file mode 100644
index 00000000..6dec3743
--- /dev/null
+++ b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_14b_instruct_128k.py
@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-14b-instruct-vllm',
+        path='Qwen/Qwen2.5-14B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=4,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]
diff --git a/opencompass/configs/models/qwen2_5/vllm_qwen2_5_32b_instruct_128k.py b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_32b_instruct_128k.py
new file mode 100644
index 00000000..5c326734
--- /dev/null
+++ b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_32b_instruct_128k.py
@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-32b-instruct-vllm',
+        path='Qwen/Qwen2.5-32B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=8,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=8),
+    )
+]
diff --git a/opencompass/configs/models/qwen2_5/vllm_qwen2_5_72b_instruct_128k.py b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_72b_instruct_128k.py
new file mode 100644
index 00000000..2a4a52fa
--- /dev/null
+++ b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_72b_instruct_128k.py
@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2_5-72b-instruct-vllm',
+        path='Qwen/Qwen2.5-72B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=8,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=8),
+    )
+]
diff --git a/opencompass/configs/models/qwen2_5/vllm_qwen2_5_7b_instruct_128k.py b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_7b_instruct_128k.py
new file mode 100644
index 00000000..db21f730
--- /dev/null
+++ b/opencompass/configs/models/qwen2_5/vllm_qwen2_5_7b_instruct_128k.py
@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-7b-instruct-vllm',
+        path='Qwen/Qwen2.5-7B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=4,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]