diff --git a/tutorials/source_en/model_infer/ms_infer/model_export.md b/tutorials/source_en/model_infer/ms_infer/model_export.md
index d7cc5ee196440c15374598ceb9fe6e0369a132f3..ef787b2a71f64d34d5f0dda9770e3f7fd4631934 100644
--- a/tutorials/source_en/model_infer/ms_infer/model_export.md
+++ b/tutorials/source_en/model_infer/ms_infer/model_export.md
@@ -20,3 +20,50 @@ ms.export(net, input_tensor, file_name='lenet', file_format='MINDIR')
 For details about the API, see [mindspore.export](https://www.mindspore.cn/docs/en/master/api_python/mindspore/mindspore.export.html?highlight=export#mindspore.export).
 
 The model export result is provided for MindSpore Lite. For details about how to use the result, see [Lite Inference Overview](../lite_infer/overview.md).
+
+## Syntax Support Scope
+
+Not all Python syntax and data types are supported for MindIR export. MindIR export has a specific support scope, and if the syntax falls outside this scope, an error will be reported during the export process.
+
+First, MindIR export only supports **strict-level graph mode**. For detailed support scope, please refer to the [Static Graph Syntax Support Documentation](https://www.mindspore.cn/tutorials/en/master/compile/static_graph.html).
+
+Second, in addition to the syntax restrictions of strict-level graph mode, MindIR has additional constraints on the types of return values. For example, returning `mindspore.dtype` is not supported. The following program will raise an error during MindIR export.
+
+```python
+import mindspore as ms
+from mindspore import nn, ops, Tensor
+
+class Model(nn.Cell):
+    def __init__(self):
+        super().__init__()
+        self.dtype = ops.DType()
+
+    def construct(self, x: Tensor) -> ms.dtype:
+        return self.dtype(x)
+```
+
+Furthermore, if a `Parameter` object is created outside of `nn.Cell`, MindIR does not support exporting that Parameter. This typically occurs in the following scenarios:
+
+- A `Parameter` is created directly in the global scope of the script.
+- A `Parameter` is created in a non `nn.Cell` class.
+- Random number generation api from the [mindspore.mint](https://www.mindspore.cn/docs/en/master/api_python/mindspore.mint.html) package are used, such as `mint.randn`, `mint.randperm`, etc., because these apis create `Parameter` objects in the global scope.
+
+For example, the following two programs will raise errors during the export process.
+
+```python
+from mindspore import Tensor, Parameter, nn
+
+param = Parameter(Tensor([1, 2, 3, 4]))  # Created outside nn.Cell
+
+class Model(nn.Cell):
+    def construct(self, x: Tensor) -> Tensor:
+        return x + param
+```
+
+```python
+from mindspore import Tensor, nn, mint
+
+class Model(nn.Cell):
+    def construct(self, n: int) -> Tensor:
+        return mint.randn(n)
+```
diff --git a/tutorials/source_zh_cn/model_infer/ms_infer/model_export.md b/tutorials/source_zh_cn/model_infer/ms_infer/model_export.md
index b85d8551982ab9a365b3bb7589a09e7317c3a95f..4d20206bea59915e934db5b8071e3b7da4d1f033 100644
--- a/tutorials/source_zh_cn/model_infer/ms_infer/model_export.md
+++ b/tutorials/source_zh_cn/model_infer/ms_infer/model_export.md
@@ -20,3 +20,50 @@ ms.export(net, input_tensor, file_name='lenet', file_format='MINDIR')
 详细接口参考链接：[mindspore.export](https://www.mindspore.cn/docs/zh-CN/master/api_python/mindspore/mindspore.export.html?highlight=export#mindspore.export)。
 
 模型导出结果提供给MindSpore Lite使用，使用方式可参考[MindSpore Lite推理](../lite_infer/overview.md)。
+
+## 语法支持范围
+
+并不是所有的 Python 语法和数据类型都支持 MindIR 导出，MindIR 导出有特定的支持范围，若不在支持范围内，导出时会报错。
+
+首先，MindIR 导出仅支持**严格图模式**，详细的支持范围，可参考[静态图语法支持](https://www.mindspore.cn/tutorials/zh-CN/master/compile/static_graph.html)。
+
+其次，除了严格图模式的语法限制，MindIR 对返回值的类型有额外约束，比如不支持返回`mindspore.dtype`。例如下面的程序，MindIR 导出的时候就会报错。
+
+```python
+import mindspore as ms
+from mindspore import nn, ops, Tensor
+
+class Model(nn.Cell):
+    def __init__(self):
+        super().__init__()
+        self.dtype = ops.DType()
+
+    def construct(self, x: Tensor) -> ms.dtype:
+        return self.dtype(x)
+```
+
+另外，如果在`nn.Cell`外创建了`Parameter`对象，MindIR 不支持导出该`Parameter`。这种情况通常发生在：
+
+- 直接在脚本的全局作用域中创建了`Parameter`。
+- 在非`nn.Cell`类中创建了`Parameter`。
+- 使用了 [mindspore.mint](https://www.mindspore.cn/docs/zh-CN/master/api_python/mindspore.mint.html) 包下的随机数生成接口，如`mint.randn`、`mint.randperm`等等，因为这些随机数接口会在全局作用域中创建`Parameter`。
+
+例如下面的两个程序，在导出过程中，就会报错。
+
+```python
+from mindspore import Tensor, Parameter, nn
+
+param = Parameter(Tensor([1, 2, 3, 4]))  # 在nn.Cell外创建
+
+class Model(nn.Cell):
+    def construct(self, x: Tensor) -> Tensor:
+        return x + param
+```
+
+```python
+from mindspore import Tensor, nn, mint
+
+class Model(nn.Cell):
+    def construct(self, n: int) -> Tensor:
+        return mint.randn(n)
+```