diff --git a/benchmarks/README.md b/benchmarks/README.md
index caff8f034214..1d715a193ea1 100644
--- a/benchmarks/README.md
+++ b/benchmarks/README.md
@@ -40,7 +40,7 @@ become available.
       <td><code>wget https://github.com/HPMLL/BurstGPT/releases/download/v1.1/BurstGPT_without_fails_2.csv</code></td>
     </tr>
     <tr>
-      <td><strong>Sonnet</strong></td>
+      <td><strong>Sonnet (deprecated)</strong></td>
       <td style="text-align: center;">✅</td>
       <td style="text-align: center;">✅</td>
       <td>Local file: <code>benchmarks/sonnet.txt</code></td>
@@ -51,6 +51,12 @@ become available.
       <td style="text-align: center;">✅</td>
       <td><code>synthetic</code></td>
     </tr>
+    <tr>
+      <td><strong>Prefix Repetition</strong></td>
+      <td style="text-align: center;">✅</td>
+      <td style="text-align: center;">✅</td>
+      <td><code>synthetic</code></td>
+    </tr>
     <tr>
       <td><strong>HuggingFace-VisionArena</strong></td>
       <td style="text-align: center;">✅</td>
@@ -592,6 +598,20 @@ python3 benchmarks/benchmark_prefix_caching.py \
   --input-length-range 128:256
 ```
 
+### Prefix Repetition Dataset
+
+```bash
+vllm bench serve \
+  --backend openai \
+  --model meta-llama/Llama-2-7b-chat-hf \
+  --dataset-name prefix_repetition \
+  --num-prompts 100 \
+  --prefix-repetition-prefix-len 512 \
+  --prefix-repetition-suffix-len 128 \
+  --prefix-repetition-num-prefixes 5 \
+  --prefix-repetition-output-len 128 
+```
+
 </details>
 
 ## ⚡ Example - Request Prioritization Benchmark
diff --git a/vllm/benchmarks/throughput.py b/vllm/benchmarks/throughput.py
index fdf6548ada5b..0c19fa6dcfdd 100644
--- a/vllm/benchmarks/throughput.py
+++ b/vllm/benchmarks/throughput.py
@@ -18,9 +18,11 @@ from transformers import (AutoModelForCausalLM, AutoTokenizer,
 
 from vllm.benchmarks.datasets import (AIMODataset, BurstGPTDataset,
                                       ConversationDataset,
-                                      InstructCoderDataset, RandomDataset,
-                                      SampleRequest, ShareGPTDataset,
-                                      SonnetDataset, VisionArenaDataset)
+                                      InstructCoderDataset,
+                                      PrefixRepetitionRandomDataset,
+                                      RandomDataset, SampleRequest,
+                                      ShareGPTDataset, SonnetDataset,
+                                      VisionArenaDataset)
 from vllm.benchmarks.lib.utils import (convert_to_pytorch_benchmark_format,
                                        write_to_json)
 from vllm.engine.arg_utils import AsyncEngineArgs, EngineArgs
@@ -327,6 +329,12 @@ def get_requests(args, tokenizer):
             dataset_cls = AIMODataset
             common_kwargs['dataset_subset'] = None
             common_kwargs['dataset_split'] = "train"
+    elif args.dataset_name == "prefix_repetition":
+        dataset_cls = PrefixRepetitionRandomDataset
+        sample_kwargs["prefix_len"] = args.prefix_repetition_prefix_len
+        sample_kwargs["suffix_len"] = args.prefix_repetition_suffix_len
+        sample_kwargs["num_prefixes"] = args.prefix_repetition_num_prefixes
+        sample_kwargs["output_len"] = args.prefix_repetition_output_len
     else:
         raise ValueError(f"Unknown dataset name: {args.dataset_name}")
     # Remove None values
@@ -356,7 +364,11 @@ def validate_args(args):
         raise ValueError(f"Unsupported backend: {args.backend}")
 
     # === Dataset Configuration ===
-    if not args.dataset and not args.dataset_path:
+    if (
+        not args.dataset
+        and not args.dataset_path
+        and args.dataset_name not in {"prefix_repetition"}
+    ):
         print(
             "When dataset path is not set, it will default to random dataset")
         args.dataset_name = 'random'
@@ -432,7 +444,10 @@ def add_cli_args(parser: argparse.ArgumentParser):
     parser.add_argument(
         "--dataset-name",
         type=str,
-        choices=["sharegpt", "random", "sonnet", "burstgpt", "hf"],
+        choices=[
+            "sharegpt", "random", "sonnet", "burstgpt", "hf",
+            "prefix_repetition"
+        ],
         help="Name of the dataset to benchmark on.",
         default="sharegpt")
     parser.add_argument(
@@ -521,6 +536,38 @@ def add_cli_args(parser: argparse.ArgumentParser):
                         default=None,
                         help="Split of the HF dataset.")
 
+    # prefix repetition dataset
+    prefix_repetition_group = parser.add_argument_group(
+        "prefix repetition dataset options")
+    prefix_repetition_group.add_argument(
+        "--prefix-repetition-prefix-len",
+        type=int,
+        default=None,
+        help="Number of prefix tokens per request, used only for prefix "
+        "repetition dataset.",
+    )
+    prefix_repetition_group.add_argument(
+        "--prefix-repetition-suffix-len",
+        type=int,
+        default=None,
+        help="Number of suffix tokens per request, used only for prefix "
+        "repetition dataset. Total input length is prefix_len + suffix_len.",
+    )
+    prefix_repetition_group.add_argument(
+        "--prefix-repetition-num-prefixes",
+        type=int,
+        default=None,
+        help="Number of prefixes to generate, used only for prefix repetition "
+        "dataset. Prompts per prefix is num_requests // num_prefixes.",
+    )
+    prefix_repetition_group.add_argument(
+        "--prefix-repetition-output-len",
+        type=int,
+        default=None,
+        help="Number of output tokens per request, used only for prefix "
+        "repetition dataset.",
+    )
+
     parser = AsyncEngineArgs.add_cli_args(parser)