diff --git a/examples/specdec_bench/run.py b/examples/specdec_bench/run.py
index f4fbf06c0e8..94932c787b8 100644
--- a/examples/specdec_bench/run.py
+++ b/examples/specdec_bench/run.py
@@ -265,7 +265,7 @@ def run_simple(args):
         type=str,
         required=False,
         default="EAGLE3",
-        choices=["EAGLE3", "EAGLE", "DRAFT_TARGET", "NGRAM", "MTP", "NONE"],
+        choices=["EAGLE3", "EAGLE", "DRAFT_TARGET", "NGRAM", "MTP", "DFLASH", "NONE"],
         help="Speculative algorithm to use",
     )
     parser.add_argument("--model_dir", type=str, required=True, help="Path to the model directory")
diff --git a/examples/specdec_bench/specdec_bench/models/sglang.py b/examples/specdec_bench/specdec_bench/models/sglang.py
index d5ff890ffd7..5e39695aff7 100644
--- a/examples/specdec_bench/specdec_bench/models/sglang.py
+++ b/examples/specdec_bench/specdec_bench/models/sglang.py
@@ -43,44 +43,48 @@ def __init__(
             speculative_algorithm = "LOOKAHEAD"
         elif speculative_algorithm == "NONE":
             speculative_algorithm = None
+
+        engine_kwargs = dict(
+            model_path=model_dir,
+            skip_tokenizer_init=True,
+            trust_remote_code=kwargs.get("trust_remote_code", False),
+            mem_fraction_static=kwargs.get("mem_fraction_static", 0.8),
+            disable_overlap_schedule=kwargs.get("disable_overlap_schedule", False),
+            tp_size=kwargs.get("tensor_parallel_size", 1),
+            ep_size=kwargs.get("moe_expert_parallel_size", 1),
+            torch_compile_max_bs=max_concurrent_requests,
+            max_running_requests=max_concurrent_requests,
+            attention_backend=kwargs.get("attention_backend"),
+            enable_torch_compile=kwargs.get("enable_torch_compile", False),
+            cuda_graph_max_bs=max_concurrent_requests,
+            disable_cuda_graph=False,
+            disable_cuda_graph_padding=True,
+        )
         if speculative_algorithm is not None:
             # https://github.com/sgl-project/sglang/pull/3582
-            self.model = sgl.Engine(
-                model_path=model_dir,
-                skip_tokenizer_init=True,
-                trust_remote_code=kwargs.get("trust_remote_code", False),
-                mem_fraction_static=0.8,
-                disable_overlap_schedule=kwargs.get("disable_overlap_schedule", False),
-                tp_size=kwargs.get("tensor_parallel_size", 1),
-                ep_size=kwargs.get("moe_expert_parallel_size", 1),
-                speculative_algorithm=speculative_algorithm,
-                speculative_num_steps=kwargs.get("speculative_num_steps", 3),
-                speculative_eagle_topk=kwargs.get("speculative_eagle_topk", 1),
-                speculative_num_draft_tokens=kwargs.get("speculative_num_draft_tokens", 4),
-                speculative_draft_model_path=kwargs.get("draft_model_dir"),
-                torch_compile_max_bs=max_concurrent_requests,
-                max_running_requests=max_concurrent_requests,
-                attention_backend=kwargs.get("attention_backend"),
-                enable_torch_compile=kwargs.get("enable_torch_compile", False),
-                cuda_graph_max_bs=max_concurrent_requests,
-                disable_cuda_graph=False,
-            )
-        else:
-            self.model = sgl.Engine(
-                model_path=model_dir,
-                skip_tokenizer_init=True,
-                trust_remote_code=kwargs.get("trust_remote_code", False),
-                mem_fraction_static=0.8,
-                disable_overlap_schedule=kwargs.get("disable_overlap_schedule", False),
-                tp_size=kwargs.get("tensor_parallel_size", 1),
-                ep_size=kwargs.get("moe_expert_parallel_size", 1),
-                torch_compile_max_bs=max_concurrent_requests,
-                max_running_requests=max_concurrent_requests,
-                attention_backend=kwargs.get("attention_backend"),
-                enable_torch_compile=kwargs.get("enable_torch_compile", False),
-                cuda_graph_max_bs=max_concurrent_requests,
-                disable_cuda_graph=False,
-            )
+            engine_kwargs["speculative_algorithm"] = speculative_algorithm
+            engine_kwargs["speculative_draft_model_path"] = kwargs.get("draft_model_dir")
+            if speculative_algorithm == "DFLASH":
+                engine_kwargs["speculative_num_draft_tokens"] = kwargs.get("speculative_num_draft_tokens", 8)
+                if "speculative_dflash_draft_window_size" in kwargs:
+                    engine_kwargs["speculative_dflash_draft_window_size"] = kwargs[
+                        "speculative_dflash_draft_window_size"
+                    ]
+                print(
+                    f"[specdec_bench] DFLASH ignores --draft_length / speculative_num_steps / "
+                    f"speculative_eagle_topk; effective draft block = "
+                    f"speculative_num_draft_tokens={engine_kwargs['speculative_num_draft_tokens']}"
+                )
+            else:
+                engine_kwargs["speculative_num_draft_tokens"] = kwargs.get("speculative_num_draft_tokens", 4)
+                engine_kwargs["speculative_num_steps"] = kwargs.get("speculative_num_steps", 3)
+                engine_kwargs["speculative_eagle_topk"] = kwargs.get("speculative_eagle_topk", 1)
+
+        # extra engine arg needed for qwen3.5
+        if "mamba_scheduler_strategy" in kwargs:
+            engine_kwargs["mamba_scheduler_strategy"] = kwargs["mamba_scheduler_strategy"]
+
+        self.model = sgl.Engine(**engine_kwargs)
 
         self.sampling_config = sampling_kwargs
 
diff --git a/examples/specdec_bench/specdec_bench/models/vllm.py b/examples/specdec_bench/specdec_bench/models/vllm.py
index 2e312e7aec8..fc595c1d579 100644
--- a/examples/specdec_bench/specdec_bench/models/vllm.py
+++ b/examples/specdec_bench/specdec_bench/models/vllm.py
@@ -63,6 +63,12 @@ def __init__(self, model_dir, max_concurrent_requests, sampling_kwargs, **kwargs
                 "method": "mtp",
                 "num_speculative_tokens": kwargs.get("speculative_num_steps", 3),
             }
+        elif kwargs.get("speculative_algorithm") == "DFLASH":
+            specdec = {
+                "method": "dflash",
+                "model": kwargs.get("draft_model_dir"),
+                "num_speculative_tokens": kwargs.get("speculative_num_draft_tokens", 8),
+            }
         elif kwargs.get("speculative_algorithm") == "NONE":
             specdec = None