diff --git a/docs/design/attention_backends.md b/docs/design/attention_backends.md
index d5d18f4ef1..edd3a93ac0 100644
--- a/docs/design/attention_backends.md
+++ b/docs/design/attention_backends.md
@@ -199,7 +199,9 @@ hardware and configuration.
 
 > **‡** Automatic selection tries FlashAttention first. On Blackwell
 > (SM100), the fallback order is TRT-LLM Ragged, FlashInfer, then
-> TokenSpeed MLA. On other GPUs, only FlashAttention is considered.
+> TokenSpeed MLA. On other GPUs, only FlashAttention is considered. TokenSpeed
+> MLA is optional; its package must use an `apache-tvm-ffi` version compatible
+> with the installed TileLang release.
 
 ### Decode Backends
 
diff --git a/requirements/cuda.txt b/requirements/cuda.txt
index 52161d05f1..ee89a13b5d 100644
--- a/requirements/cuda.txt
+++ b/requirements/cuda.txt
@@ -24,8 +24,8 @@ fastsafetensors >= 0.3.3
 nvidia-cutlass-dsl[cu13]==4.7.0
 quack-kernels>=0.3.3
 
-# Tokenspeed_MLA for faster mla with spec decode
-tokenspeed-mla==0.2.5
+# TokenSpeed MLA is optional. Version 0.2.5 pins apache-tvm-ffi==0.1.13,
+# which cannot coexist with the TileLang version pinned above.
 
 # Humming kernels for quantization gemm
 humming-kernels[cu13]==0.1.2
diff --git a/vllm/v1/attention/backends/mla/prefill/tokenspeed_mla.py b/vllm/v1/attention/backends/mla/prefill/tokenspeed_mla.py
index d6e4fca172..477bed24b7 100644
--- a/vllm/v1/attention/backends/mla/prefill/tokenspeed_mla.py
+++ b/vllm/v1/attention/backends/mla/prefill/tokenspeed_mla.py
@@ -30,8 +30,9 @@ def supports_compute_capability(cls, device_capability: "DeviceCapability") -> b
         return device_capability.major == 10
 
     _INSTALL_HINT = (
-        "tokenspeed_mla package is not installed. "
-        "Install it with: `uv pip install tokenspeed-mla`"
+        "tokenspeed_mla is an optional package and is not installed. "
+        "Its apache-tvm-ffi requirement must be compatible with the installed "
+        "TileLang version."
     )
 
     @classmethod
diff --git a/vllm/v1/attention/backends/mla/tokenspeed_mla.py b/vllm/v1/attention/backends/mla/tokenspeed_mla.py
index 6c8dedd77f..2af65cb9ce 100644
--- a/vllm/v1/attention/backends/mla/tokenspeed_mla.py
+++ b/vllm/v1/attention/backends/mla/tokenspeed_mla.py
@@ -101,8 +101,9 @@ def supports_combination(
             import tokenspeed_mla  # noqa: F401
         except ImportError:
             return (
-                "tokenspeed_mla package is not installed. "
-                "Install it with: `uv pip install tokenspeed-mla`"
+                "tokenspeed_mla is an optional package and is not installed. "
+                "Its apache-tvm-ffi requirement must be compatible with the "
+                "installed TileLang version."
             )
 
         # tokenspeed_mla CuTe DSL kernel is shape-specialized for DeepSeek R1