diff --git a/vllm-eval-harness/configs/deepseek-ai/DeepSeek-R1.yml b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-R1.yml
new file mode 100644
index 00000000..c97b4227
--- /dev/null
+++ b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-R1.yml
@@ -0,0 +1,14 @@
+model_name: "deepseek-ai/DeepSeek-R1"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.1.yml b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.1.yml
new file mode 100644
index 00000000..833fae02
--- /dev/null
+++ b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.1.yml
@@ -0,0 +1,14 @@
+model_name: "deepseek-ai/DeepSeek-V3.1"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.2-Exp.yml b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.2-Exp.yml
new file mode 100644
index 00000000..38543069
--- /dev/null
+++ b/vllm-eval-harness/configs/deepseek-ai/DeepSeek-V3.2-Exp.yml
@@ -0,0 +1,14 @@
+model_name: "deepseek-ai/DeepSeek-V3.2-Exp"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/google/gemma-3-27b-it.yml b/vllm-eval-harness/configs/google/gemma-3-27b-it.yml
new file mode 100644
index 00000000..a9f51d88
--- /dev/null
+++ b/vllm-eval-harness/configs/google/gemma-3-27b-it.yml
@@ -0,0 +1,14 @@
+model_name: "google/gemma-3-27b-it"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8.yml b/vllm-eval-harness/configs/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8.yml
new file mode 100644
index 00000000..4279893c
--- /dev/null
+++ b/vllm-eval-harness/configs/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8.yml
@@ -0,0 +1,14 @@
+model_name: "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/meta-llama/Llama-4-Scout-17B-16E-Instruct.yml b/vllm-eval-harness/configs/meta-llama/Llama-4-Scout-17B-16E-Instruct.yml
new file mode 100644
index 00000000..965486b2
--- /dev/null
+++ b/vllm-eval-harness/configs/meta-llama/Llama-4-Scout-17B-16E-Instruct.yml
@@ -0,0 +1,14 @@
+model_name: "meta-llama/Llama-4-Scout-17B-16E-Instruct"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 4
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/openai/gpt-oss-120b.yml b/vllm-eval-harness/configs/openai/gpt-oss-120b.yml
new file mode 100644
index 00000000..3b328220
--- /dev/null
+++ b/vllm-eval-harness/configs/openai/gpt-oss-120b.yml
@@ -0,0 +1,14 @@
+model_name: "openai/gpt-oss-120b"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/openai/gpt-oss-20b.yml b/vllm-eval-harness/configs/openai/gpt-oss-20b.yml
new file mode 100644
index 00000000..16f01bf9
--- /dev/null
+++ b/vllm-eval-harness/configs/openai/gpt-oss-20b.yml
@@ -0,0 +1,14 @@
+model_name: "openai/gpt-oss-20b"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 1
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/qwen/Qwen3-30B-A3B.yml b/vllm-eval-harness/configs/qwen/Qwen3-30B-A3B.yml
new file mode 100644
index 00000000..d8c8a6ff
--- /dev/null
+++ b/vllm-eval-harness/configs/qwen/Qwen3-30B-A3B.yml
@@ -0,0 +1,14 @@
+model_name: "Qwen/Qwen3-30B-A3B"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 8
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/configs/qwen/Qwen3-8B.yml b/vllm-eval-harness/configs/qwen/Qwen3-8B.yml
new file mode 100644
index 00000000..18ac330a
--- /dev/null
+++ b/vllm-eval-harness/configs/qwen/Qwen3-8B.yml
@@ -0,0 +1,14 @@
+model_name: "Qwen/Qwen3-8B"
+tasks:
+- name: "gsm8k"
+  device: b200
+  tp: 1
+  # Adopted from vLLM lm-eval-harness, set the value to 0 if there is no baseline
+  metrics:
+    - name: "exact_match,strict-match"
+      value: 0
+    - name: "exact_match,flexible-extract"
+      value: 0
+limit: 1000
+num_fewshot: 5
+trust_remote_code: True
diff --git a/vllm-eval-harness/run_vllm_eval_harness.py b/vllm-eval-harness/run_vllm_eval_harness.py
new file mode 100644
index 00000000..3a2de086
--- /dev/null
+++ b/vllm-eval-harness/run_vllm_eval_harness.py
@@ -0,0 +1,186 @@
+import json
+import os
+import glob
+import lm_eval
+import yaml
+from logging import warning, info
+from argparse import Action, ArgumentParser, Namespace
+import torch
+from typing import Dict, Any, List, Optional
+
+
+# See lm-eval docs for the list of acceptable values
+LM_EVAL_MODEL_SOURCE = os.environ.get("LM_EVAL_MODEL_SOURCE", "vllm")
+
+
+class ValidateDir(Action):
+    def __call__(
+        self,
+        parser: ArgumentParser,
+        namespace: Namespace,
+        values: Any,
+        option_string: Optional[str] = None,
+    ) -> None:
+        if os.path.isdir(values):
+            setattr(namespace, self.dest, values)
+            return
+
+        parser.error(f"{values} is not a valid directory")
+
+
+def parse_args() -> Any:
+    parser = ArgumentParser("Run vLLM lm-eval harness")
+
+    parser.add_argument(
+        "--configs-dir",
+        type=str,
+        action=ValidateDir,
+        help="the directory contains vLLM lm-eval harness configs",
+        required=True,
+    )
+    parser.add_argument(
+        "--models",
+        type=str,
+        default="",
+        help="the comma-separated list of models to evaluate (optional)",
+    )
+    parser.add_argument(
+        "--tasks",
+        type=str,
+        default="",
+        help="the comma-separated list of tasks to evaluate (optional)",
+    )
+
+    return parser.parse_args()
+
+
+def convert_to_pytorch_benchmark_format(
+    model_name: str, tp_size: int, results: Dict[str, Any]
+) -> List[Any]:
+    records = []
+    configs = results.get("configs", {})
+
+    for task_name, metrics in results.get("results", {}).items():
+        for metric_name, metric_value in metrics.items():
+            if type(metric_value) is str:
+                continue
+
+            record = {
+                "benchmark": {
+                    "name": "vLLM lm-eval harness",
+                    "extra_info": {
+                        "args": {
+                            "tensor_parallel_size": tp_size,
+                        },
+                        "configs": configs.get(task_name, {}),
+                    },
+                },
+                "model": {
+                    "name": model_name,
+                },
+                "metric": {
+                    "name": metric_name,
+                    "benchmark_values": [metric_value],
+                },
+            }
+            records.append(record)
+
+    return records
+
+
+def run(
+    model_name: str, tasks: List[str], tp_size: int, config: Dict[str, Any]
+) -> Dict[str, Any]:
+    trust_remote_code = config.get("trust_remote_code", False)
+    max_model_len = config.get("max_model_len", "auto")
+
+    model_args = (
+        f"pretrained={model_name},"
+        f"tensor_parallel_size={tp_size},"
+        f"add_bos_token=true,"
+        f"trust_remote_code={trust_remote_code},"
+        f"max_model_len={max_model_len}"
+    )
+    info(f"Evaluating {model_name} with {model_args}")
+    return lm_eval.simple_evaluate(
+        model=LM_EVAL_MODEL_SOURCE,
+        model_args=model_args,
+        tasks=tasks,
+        num_fewshot=config["num_fewshot"],
+        limit=config["limit"],
+        batch_size="auto",
+    )
+
+
+def run_lm_eval(configs_dir: str, models: List[str], tasks: List[str]) -> None:
+    device_name = torch.cuda.get_device_name().lower()
+    device_count = torch.cuda.device_count()
+
+    results_dir = os.path.join(configs_dir, "results")
+    os.makedirs(results_dir, exist_ok=True)
+
+    for file in glob.glob(f"{configs_dir}/**/*.yml", recursive=True):
+        with open(file) as f:
+            config = yaml.safe_load(f)
+        # Check the model name
+        model_name = config.get("model_name", "").lower()
+        if models and model_name not in models:
+            info(f"Skip {model_name} from {file}")
+            continue
+
+        tp_size = 0
+        selected_tasks = []
+
+        # Check the lm-eval tasks, the selected device, and tp
+        for t in config.get("tasks", []):
+            task_name = t["name"]
+            if not task_name:
+                warning(f"{model_name} from {file}: skip missing task")
+                continue
+
+            if tasks and task_name not in tasks:
+                info(f"{model_name} from {file}: {task_name} not selected")
+
+            selected_device = t["device"].lower()
+            if selected_device not in device_name:
+                continue
+
+            tp = t["tp"]
+            if device_count < tp:
+                warning(
+                    f"{model_name} from {file}: device count {device_count} < tp {tp} in {task_name}"
+                )
+                continue
+
+            selected_tasks.append(task_name)
+            if not tp_size:
+                tp_size = tp
+            assert tp_size == tp
+
+        if not selected_tasks:
+            info(f"Skip {model_name} from {file}: no task")
+            continue
+
+        results = run(model_name, selected_tasks, tp_size, config)
+        results_pytorch_format = convert_to_pytorch_benchmark_format(
+            model_name, tp_size, results
+        )
+
+        results_file = os.path.splitext(os.path.basename(file))[0]
+        # Dump the results from lm-eval
+        with open(os.path.join(results_dir, f"{results_file}_lm_eval.json"), "w") as f:
+            json.dump(results, f, indent=2)
+        # Dump the results that can be uploaded to PyTorch OSS benchmark infra
+        with open(os.path.join(results_dir, f"{results_file}_pytorch.json"), "w") as f:
+            json.dump(results_pytorch_format, f, indent=2)
+
+
+def main() -> None:
+    args = parse_args()
+    models = [m.strip().lower() for m in args.models.split(",") if m.strip()]
+    tasks = [m.strip().lower() for m in args.tasks.split(",") if m.strip()]
+    run_lm_eval(args.configs_dir, models, tasks)
+
+
+if __name__ == "__main__":
+    main()