diff --git a/docs/OperatorKernels.md b/docs/OperatorKernels.md
index 507c6722bc349..01d91269622fc 100644
--- a/docs/OperatorKernels.md
+++ b/docs/OperatorKernels.md
@@ -940,7 +940,9 @@ Do not modify directly.*
 |||[2, 10]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |Sqrt|*in* X:**T**<br> *out* Y:**T**|13+|**T** = tensor(bfloat16), tensor(double), tensor(float), tensor(float16)|
 |||[6, 12]|**T** = tensor(double), tensor(float), tensor(float16)|
-|Squeeze|*in* data:**T**<br> *in* axes:**tensor(int64)**<br> *out* squeezed:**T**<br><br>or<br><br>*in* data:**T**<br> *out* squeezed:**T**|23+|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|Squeeze|*in* data:**T**<br> *in* axes:**tensor(int64)**<br> *out* squeezed:**T**<br><br>or<br><br>*in* data:**T**<br> *out* squeezed:**T**|25+|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|||24|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|||23|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[21, 22]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[13, 20]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[11, 12]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
@@ -966,7 +968,9 @@ Do not modify directly.*
 |||[13, 20]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[1, 12]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |Trilu|*in* input:**T**<br> *in* k:**tensor(int64)**<br> *out* output:**T**|14+|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
-|Unsqueeze|*in* data:**T**<br> *in* axes:**tensor(int64)**<br> *out* expanded:**T**<br><br>or<br><br>*in* data:**T**<br> *out* expanded:**T**|23+|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|Unsqueeze|*in* data:**T**<br> *in* axes:**tensor(int64)**<br> *out* expanded:**T**<br><br>or<br><br>*in* data:**T**<br> *out* expanded:**T**|25+|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|||24|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
+|||23|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[21, 22]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[13, 20]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
 |||[11, 12]|**T** = tensor(bfloat16), tensor(bool), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)|
diff --git a/onnxruntime/core/providers/cuda/cuda_execution_provider.cc b/onnxruntime/core/providers/cuda/cuda_execution_provider.cc
index 60ac16018f539..4dbc1edeb9b91 100755
--- a/onnxruntime/core/providers/cuda/cuda_execution_provider.cc
+++ b/onnxruntime/core/providers/cuda/cuda_execution_provider.cc
@@ -1626,8 +1626,8 @@ class ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain,
 class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Shape);
 class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Reshape);
 class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Transpose);
-class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Squeeze);
-class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Unsqueeze);
+class ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, 23, Squeeze);
+class ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, 23, Unsqueeze);
 
 #if !defined(DISABLE_FLOAT4_TYPES)
 class ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Float4E2M1x2, Cast);
@@ -1638,6 +1638,12 @@ class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, T
 class ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, float, Attention);
 class ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, MLFloat16, Attention);
 class ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, BFloat16, Attention);
+class ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, 24, Squeeze);
+class ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, 24, Unsqueeze);
+
+// Opset 25.
+class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 25, Squeeze);
+class ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 25, Unsqueeze);
 
 #endif
 
@@ -2713,14 +2719,20 @@ static Status RegisterCudaKernels(KernelRegistry& kernel_registry) {
       BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Shape)>,
       BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Reshape)>,
       BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Transpose)>,
-      BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Squeeze)>,
-      BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, Unsqueeze)>,
+      BuildKernelCreateInfo<ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, 23, Squeeze)>,
+      BuildKernelCreateInfo<ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 23, 23, Unsqueeze)>,
 
       // Opset 24
       BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, TensorScatter)>,
       BuildKernelCreateInfo<ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, float, Attention)>,
       BuildKernelCreateInfo<ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, MLFloat16, Attention)>,
       BuildKernelCreateInfo<ONNX_OPERATOR_TYPED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, BFloat16, Attention)>,
+      BuildKernelCreateInfo<ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, 24, Squeeze)>,
+      BuildKernelCreateInfo<ONNX_OPERATOR_VERSIONED_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 24, 24, Unsqueeze)>,
+
+      // Opset 25
+      BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 25, Squeeze)>,
+      BuildKernelCreateInfo<ONNX_OPERATOR_KERNEL_CLASS_NAME(kCudaExecutionProvider, kOnnxDomain, 25, Unsqueeze)>,
 #endif
   };
 
diff --git a/onnxruntime/core/providers/cuda/tensor/squeeze.cc b/onnxruntime/core/providers/cuda/tensor/squeeze.cc
index 3ccc57acf2252..6c32d7356850b 100644
--- a/onnxruntime/core/providers/cuda/tensor/squeeze.cc
+++ b/onnxruntime/core/providers/cuda/tensor/squeeze.cc
@@ -50,10 +50,32 @@ ONNX_OPERATOR_VERSIONED_KERNEL_EX(
         .InputMemoryType(OrtMemTypeCPUInput, 1),
     Squeeze);
 
+ONNX_OPERATOR_VERSIONED_KERNEL_EX(
+    Squeeze,
+    kOnnxDomain,
+    23, 23,
+    kCudaExecutionProvider,
+    (*KernelDefBuilder::Create())
+        .Alias(0, 0)
+        .TypeConstraint("T", DataTypeImpl::AllFixedSizeTensorTypes())
+        .InputMemoryType(OrtMemTypeCPUInput, 1),
+    Squeeze);
+
+ONNX_OPERATOR_VERSIONED_KERNEL_EX(
+    Squeeze,
+    kOnnxDomain,
+    24, 24,
+    kCudaExecutionProvider,
+    (*KernelDefBuilder::Create())
+        .Alias(0, 0)
+        .TypeConstraint("T", DataTypeImpl::AllFixedSizeTensorTypes())
+        .InputMemoryType(OrtMemTypeCPUInput, 1),
+    Squeeze);
+
 ONNX_OPERATOR_KERNEL_EX(
     Squeeze,
     kOnnxDomain,
-    23,
+    25,
     kCudaExecutionProvider,
     (*KernelDefBuilder::Create())
         .Alias(0, 0)
diff --git a/onnxruntime/core/providers/cuda/tensor/unsqueeze.cc b/onnxruntime/core/providers/cuda/tensor/unsqueeze.cc
index 8d5a601aadd6e..411f94f31c7ba 100644
--- a/onnxruntime/core/providers/cuda/tensor/unsqueeze.cc
+++ b/onnxruntime/core/providers/cuda/tensor/unsqueeze.cc
@@ -50,10 +50,32 @@ ONNX_OPERATOR_VERSIONED_KERNEL_EX(
         .InputMemoryType(OrtMemTypeCPUInput, 1),
     Unsqueeze);
 
+ONNX_OPERATOR_VERSIONED_KERNEL_EX(
+    Unsqueeze,
+    kOnnxDomain,
+    23, 23,
+    kCudaExecutionProvider,
+    (*KernelDefBuilder::Create())
+        .Alias(0, 0)
+        .TypeConstraint("T", DataTypeImpl::AllFixedSizeTensorTypes())
+        .InputMemoryType(OrtMemTypeCPUInput, 1),
+    Unsqueeze);
+
+ONNX_OPERATOR_VERSIONED_KERNEL_EX(
+    Unsqueeze,
+    kOnnxDomain,
+    24, 24,
+    kCudaExecutionProvider,
+    (*KernelDefBuilder::Create())
+        .Alias(0, 0)
+        .TypeConstraint("T", DataTypeImpl::AllFixedSizeTensorTypes())
+        .InputMemoryType(OrtMemTypeCPUInput, 1),
+    Unsqueeze);
+
 ONNX_OPERATOR_KERNEL_EX(
     Unsqueeze,
     kOnnxDomain,
-    23,
+    25,
     kCudaExecutionProvider,
     (*KernelDefBuilder::Create())
         .Alias(0, 0)