diff --git a/functorch/csrc/BatchRulesBinaryOps.cpp b/functorch/csrc/BatchRulesBinaryOps.cpp
index 2f5e9ae305..7367f43a77 100644
--- a/functorch/csrc/BatchRulesBinaryOps.cpp
+++ b/functorch/csrc/BatchRulesBinaryOps.cpp
@@ -183,6 +183,9 @@ TORCH_LIBRARY_IMPL(aten, FT_BATCHED_KEY, m) {
   UNARY_POINTWISE(clamp_max);
   POINTWISE_BOXED(clamp_max_);
 
+  UNARY_POINTWISE(clip);
+  POINTWISE_BOXED(clip.Tensor);
+
   // Commented out so we have a test op
   // BINARY_SCALAR_2(copysign, Tensor, Scalar);
   BINARY_SCALAR_2(div, Tensor, Scalar);
diff --git a/functorch/csrc/BatchingRegistrations.cpp b/functorch/csrc/BatchingRegistrations.cpp
index 02f6cef3b0..1b92543a70 100644
--- a/functorch/csrc/BatchingRegistrations.cpp
+++ b/functorch/csrc/BatchingRegistrations.cpp
@@ -888,7 +888,6 @@ TORCH_LIBRARY_IMPL(aten, FT_BATCHED_KEY, m) {
   m.impl("view_as", native::view_as); // composite wrt autograd
 
   m.impl("addmm", addmm_batching_rule);
-  // clamp operations
 
 // unary pointwise, out-of-place, no additional arguments.
 #define TO_BATCHING_RULE(name, ...) \