diff --git a/pkg/model/field.go b/pkg/model/field.go
index deb3f4b3..ab7f3ab6 100644
--- a/pkg/model/field.go
+++ b/pkg/model/field.go
@@ -92,6 +92,41 @@ type Field struct {
 //	// please note that this field is updated on the service
 //	// side"
 func (f *Field) GetDocumentation() string {
+	return indentDocComment(f.buildDocumentation())
+}
+
+// docCommentIndent is prefixed to every line of a generated field doc comment
+// so the comment begins past column 1. gofmt only re-flows a doc comment
+// (normalizing list markers/indentation via go/doc/comment) when the comment
+// starts at column 1; an already-indented comment is left as written. Emitting
+// the comment pre-indented keeps generated output identical across Go
+// toolchains -- Go <=1.26 reformatted column-1 field comments while Go 1.27+
+// does not (see go/printer commit c1f0b9b) -- without the generator having to
+// reproduce gofmt's canonical list formatting itself. gofmt still re-derives
+// the actual struct indentation, so a single tab here is sufficient regardless
+// of nesting depth.
+const docCommentIndent = "\t"
+
+// indentDocComment prefixes docCommentIndent to every non-empty comment line so
+// the whole comment (SDK docs, appended Regex Pattern, and any user-provided
+// prepend/append text) is indented uniformly.
+func indentDocComment(doc string) string {
+	if doc == "" {
+		return ""
+	}
+	lines := strings.Split(doc, "\n")
+	for i, line := range lines {
+		if line != "" {
+			lines[i] = docCommentIndent + line
+		}
+	}
+	return strings.Join(lines, "\n")
+}
+
+// buildDocumentation assembles the field's raw (unindented) doc comment from
+// the SDK documentation, an optional Regex Pattern line, and any user-provided
+// documentation.yaml override/append/prepend text.
+func (f *Field) buildDocumentation() string {
 	cfg := f.GetFieldDocsConfig()
 
 	hasShapeDoc := false
diff --git a/pkg/model/field_test.go b/pkg/model/field_test.go
index 0fdacac7..eed54a78 100644
--- a/pkg/model/field_test.go
+++ b/pkg/model/field_test.go
@@ -40,7 +40,7 @@ func TestFieldDocumentation(t *testing.T) {
 	require.NotNil(ltdField.ShapeRef)
 
 	require.Equal(
-		"// The desired Kubernetes version for your cluster. If you don't specify a value\n// here, the default version available in Amazon EKS is used.\n// \n// The default version might not be the latest version available.",
+		"\t// The desired Kubernetes version for your cluster. If you don't specify a value\n\t// here, the default version available in Amazon EKS is used.\n\t// \n\t// The default version might not be the latest version available.",
 		ltdField.GetDocumentation(),
 	)
 
@@ -49,7 +49,7 @@ func TestFieldDocumentation(t *testing.T) {
 	require.NotNil(prependField.ShapeRef)
 
 	require.True(
-		strings.HasPrefix(prependField.GetDocumentation(), "// !!! Let's take it from the top"),
+		strings.HasPrefix(prependField.GetDocumentation(), "\t// !!! Let's take it from the top"),
 	)
 
 	appendField := crd.Fields["RoleARN"]
@@ -65,9 +65,9 @@ func TestFieldDocumentation(t *testing.T) {
 	require.NotNil(overrideField.ShapeRef)
 
 	require.Equal(
-		"// !!! All your docs has become mine\n"+
-			"// \n"+
-			"// That whitespace is entirely on purpose",
+		"\t// !!! All your docs has become mine\n"+
+			"\t// \n"+
+			"\t// That whitespace is entirely on purpose",
 		overrideField.GetDocumentation(),
 	)
 
@@ -398,7 +398,7 @@ func TestFieldWithPattern(t *testing.T) {
 	require.NotEmpty(ltdField.ShapeRef.Shape.Pattern)
 
 	require.Equal(
-		"// The name of your cluster.\n//\n// Regex Pattern: `^[0-9A-Za-z][A-Za-z0-9\\-_]*$`",
+		"\t// The name of your cluster.\n\t//\n\t// Regex Pattern: `^[0-9A-Za-z][A-Za-z0-9\\-_]*$`",
 		ltdField.GetDocumentation(),
 	)
 }