diff --git a/stubgen/generator.py b/stubgen/generator.py
new file mode 100644
index 0000000..18613d9
--- /dev/null
+++ b/stubgen/generator.py
@@ -0,0 +1,449 @@
+"""Stub generator module for creating .pyi files."""
+
+import ast
+import re
+import sys
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set
+
+from stubgen.parser import FileInfo, Function, Class, Variable, Import
+from stubgen.inferrer import Inferrer
+
+
+def _split_type_args(s: str) -> List[str]:
+    """Split type arguments handling nested brackets."""
+    result = []
+    current = ""
+    depth = 0
+
+    for char in s:
+        if char == '[':
+            depth += 1
+            current += char
+        elif char == ']':
+            depth -= 1
+            current += char
+        elif char == ',' and depth == 0:
+            if current.strip():
+                result.append(current.strip())
+            current = ""
+        else:
+            current += char
+
+    if current.strip():
+        result.append(current.strip())
+
+    return result
+
+
+class StubWriter:
+    """Writes .pyi stub files from parsed and inferred information."""
+    
+    def __init__(self, inferrer: Optional[Inferrer] = None):
+        self.inferrer = inferrer or Inferrer()
+        self.imports: Set[str] = set()
+    
+    def generate(self, file_info: FileInfo) -> str:
+        """Generate a complete .pyi stub content."""
+        self.imports = set()
+        lines = []
+        
+        if file_info.docstring:
+            lines.append(f'"""{file_info.docstring}"""')
+            lines.append("")
+        
+        self._process_imports(file_info.imports)
+        
+        if self.imports:
+            lines.extend(sorted(self.imports))
+            lines.append("")
+        
+        for var in file_info.variables:
+            var_lines = self._format_variable(var)
+            lines.extend(var_lines)
+        
+        if file_info.variables:
+            lines.append("")
+        
+        for cls in file_info.classes:
+            class_lines = self._format_class(cls)
+            lines.extend(class_lines)
+            lines.append("")
+        
+        for func in file_info.functions:
+            func_lines = self._format_function(func)
+            lines.extend(func_lines)
+            lines.append("")
+        
+        return "\n".join(lines)
+    
+    def _process_imports(self, imports: List[Import]):
+        for imp in imports:
+            if imp.is_from:
+                prefix = "." * imp.level
+                if imp.module:
+                    prefix += imp.module
+                if imp.names:
+                    parts = []
+                    for name in imp.names:
+                        if name in imp.as_names:
+                            parts.append(f"{name} as {imp.as_names[name]}")
+                        else:
+                            parts.append(name)
+                    import_str = f"from {prefix} import {', '.join(parts)}"
+                    self.imports.add(import_str)
+            else:
+                if imp.module and imp.names:
+                    for name in imp.names:
+                        if name in imp.as_names:
+                            self.imports.add(f"import {imp.module} as {imp.as_names[name]}")
+                        else:
+                            self.imports.add(f"import {imp.module}.{name}")
+                elif imp.module:
+                    self.imports.add(f"import {imp.module}")
+                elif imp.names:
+                    for name in imp.names:
+                        if name in imp.as_names:
+                            self.imports.add(f"import {name} as {imp.as_names[name]}")
+                        else:
+                            self.imports.add(f"import {name}")
+    
+    def _format_variable(self, var: Variable) -> List[str]:
+        lines = []
+        
+        if var.annotation:
+            annotation_str = var.annotation.to_string()
+            self._add_needed_import(annotation_str)
+            lines.append(f"{var.name}: {annotation_str}")
+        else:
+            lines.append(f"{var.name}: Any")
+        
+        return lines
+    
+    def _format_function(self, func: Function, inferred_info: Dict = None) -> List[str]:
+        lines = []
+        
+        for decorator in func.decorators:
+            lines.append(f"@{decorator}")
+        
+        async_prefix = "async " if func.is_async else ""
+        
+        args_parts = []
+        
+        defaults_offset = len(func.args.args) - len(func.args.defaults)
+        
+        for i, arg in enumerate(func.args.args):
+            arg_str = arg.arg
+            
+            if arg.annotation:
+                ann_str = ast.unparse(arg.annotation)
+                self._add_needed_import(ann_str)
+                arg_str += f": {ann_str}"
+            elif inferred_info and arg.arg in inferred_info.get("args", {}):
+                arg_info = inferred_info["args"][arg.arg]
+                arg_type = arg_info.get("type", "Any")
+                self._add_needed_import(arg_type)
+                if "None" in str(arg_type):
+                    arg_str += f": {arg_type}"
+            elif i >= defaults_offset and i - defaults_offset < len(func.args.defaults):
+                default = func.args.defaults[i - defaults_offset]
+                default_type = self._infer_default_type(default)
+                if default_type != "Any":
+                    self._add_needed_import(default_type)
+                    arg_str += f": {default_type}"
+            
+            args_parts.append(arg_str)
+        
+        if func.args.vararg:
+            vararg_type = "tuple[Any, ...]"
+            self._add_needed_import(vararg_type)
+            args_parts.append(f"*{func.args.vararg.arg}: {vararg_type}")
+        
+        for kwonly in func.args.kwonlyargs:
+            kw_str = kwonly.arg
+            if kwonly.annotation:
+                kw_str += f": {ast.unparse(kwonly.annotation)}"
+            args_parts.append(kw_str)
+        
+        if func.args.kwarg:
+            kwarg_type = "dict[str, Any]"
+            self._add_needed_import(kwarg_type)
+            args_parts.append(f"**{func.args.kwarg.arg}: {kwarg_type}")
+        
+        args_str = ", ".join(args_parts)
+        
+        return_type = "None"
+        if func.returns:
+            if hasattr(func.returns, 'annotation') and func.returns.annotation:
+                return_type = ast.unparse(func.returns.annotation)
+            else:
+                return_type = ast.unparse(func.returns)
+            self._add_needed_import(return_type)
+        elif inferred_info:
+            ret_info = inferred_info.get("returns", {})
+            return_type = ret_info.get("type", "None")
+            self._add_needed_import(return_type)
+        
+        lines.append(f"{async_prefix}def {func.name}({args_str}) -> {return_type}: ...")
+        
+        return lines
+    
+    def _format_class(self, cls: Class, inferred_info: Dict = None) -> List[str]:
+        lines = []
+        
+        for decorator in cls.decorators:
+            lines.append(f"@{decorator}")
+        
+        bases = ", ".join(cls.bases) if cls.bases else "object"
+        lines.append(f"class {cls.name}({bases}):")
+        
+        indent = "    "
+        
+        for attr in cls.attributes:
+            if attr.annotation:
+                attr_type = attr.annotation.to_string()
+                self._add_needed_import(attr_type)
+                lines.append(f"{indent}{attr.name}: {attr_type}")
+            else:
+                lines.append(f"{indent}{attr.name}: Any")
+        
+        if cls.attributes and cls.methods:
+            lines.append("")
+        
+        if inferred_info:
+            methods = inferred_info.get("methods", {})
+        else:
+            methods = {}
+        
+        for method in cls.methods:
+            method_inferred = methods.get(method.name, {})
+            method_lines = self._format_function(method, method_inferred)
+            for line in method_lines:
+                lines.append(f"{indent}{line}")
+        
+        for nested_cls in cls.body:
+            if isinstance(nested_cls, ast.ClassDef):
+                nested_info = inferred_info.get("classes", {}).get(nested_cls.name, {}) if inferred_info else {}
+                for line in lines:
+                    pass
+                lines.append(f"{indent}class {nested_cls.name}:")
+                lines.append(f"{indent}    pass")
+        
+        if not cls.attributes and not cls.methods and not any(isinstance(n, ast.ClassDef) for n in cls.body):
+            lines.append(f"{indent}pass")
+        
+        return lines
+    
+    def _infer_default_type(self, default: ast.expr) -> str:
+        if default is None:
+            return "Any"
+        
+        if isinstance(default, ast.Constant):
+            value = default.value
+            if value is None:
+                return "None"
+            elif isinstance(value, bool):
+                return "bool"
+            elif isinstance(value, int):
+                return "int"
+            elif isinstance(value, float):
+                return "float"
+            elif isinstance(value, str):
+                return "str"
+        elif isinstance(default, ast.List):
+            return "list"
+        elif isinstance(default, ast.Dict):
+            return "dict"
+        elif isinstance(default, ast.Tuple):
+            return "tuple"
+        elif isinstance(default, ast.Set):
+            return "set"
+        
+        return "Any"
+    
+    def _add_needed_import(self, type_str: str):
+        """Add import statements for types that need them."""
+        if not type_str or type_str == "Any":
+            return
+        
+        type_str = type_str.strip()
+        
+        container_match = re.match(r'(List|Dict|Tuple|Set|FrozenSet|Callable|Type|Union|Optional)\[', type_str)
+        if container_match:
+            container_type = container_match.group(1)
+            if container_type == "List":
+                self.imports.add("from typing import List")
+            elif container_type == "Dict":
+                self.imports.add("from typing import Dict")
+            elif container_type == "Tuple":
+                self.imports.add("from typing import Tuple")
+            elif container_type == "Set":
+                self.imports.add("from typing import Set")
+            elif container_type == "FrozenSet":
+                self.imports.add("from typing import FrozenSet")
+            elif container_type == "Callable":
+                self.imports.add("from typing import Callable")
+            elif container_type == "Type":
+                self.imports.add("from typing import Type")
+            elif container_type in ("Union", "Optional"):
+                self.imports.add("from typing import Union, Optional")
+            
+            inner_types = type_str[container_match.end():-1]
+            inner_type_list = _split_type_args(inner_types)
+            for inner in inner_type_list:
+                self._add_needed_import(inner)
+        
+        generic_match = re.match(r'([A-Z][a-zA-Z0-9_]*)\[', type_str)
+        if generic_match:
+            type_name = generic_match.group(1)
+            builtin_types = {"str", "int", "float", "bool", "list", "dict", "set", "tuple", "type", "object"}
+            if type_name not in builtin_types:
+                self.imports.add(f"from typing import {type_name}")
+        
+        self.imports.add("from typing import Any")
+
+
+class StubGenerator:
+    """Main stub generator class that orchestrates parsing and generation."""
+
+    def __init__(self, inferrer: Optional[Inferrer] = None, interactive: bool = False):
+        self.inferrer = inferrer or Inferrer()
+        self.writer = StubWriter(self.inferrer)
+        self.interactive = interactive
+
+    def confirm_types_interactively(self, inferred_types: Dict[str, Any]) -> Dict[str, Any]:
+        """Allow user to confirm or modify inferred types."""
+        import click
+
+        confirmed_types = inferred_types.copy()
+
+        if not inferred_types:
+            return confirmed_types
+
+        for func_name, type_info in inferred_types.get("functions", {}).items():
+            click.echo(f"\nFunction: {func_name}")
+
+            args_info = type_info.get("args", {})
+            for arg_name, arg_info in args_info.items():
+                if arg_info.get("source") != "annotation":
+                    current_type = arg_info.get("type", "Any")
+                    confidence = arg_info.get("confidence", 0.5)
+                    click.echo(f"  {arg_name}: {current_type} (confidence: {confidence})")
+
+            returns_info = type_info.get("returns", {})
+            if returns_info.get("source") != "annotation":
+                current_type = returns_info.get("type", "None")
+                confidence = returns_info.get("confidence", 0.5)
+                click.echo(f"  return: {current_type} (confidence: {confidence})")
+
+        return confirmed_types
+
+    def generate(self, tree, inferred_types: Optional[Dict[str, Any]] = None) -> str:
+        """Generate stub content from AST tree and inferred types."""
+        from stubgen.parser import FileInfo
+
+        file_info = FileInfo(path=Path("."))
+        file_info = self._extract_info_from_tree(tree, file_info)
+
+        return self.writer.generate(file_info)
+
+    def _extract_info_from_tree(self, tree, file_info: FileInfo) -> FileInfo:
+        """Extract file info from AST tree."""
+        from stubgen.parser import Function, Class, Variable, TypeHint
+        import ast
+
+        for node in ast.walk(tree):
+            if isinstance(node, ast.FunctionDef) or isinstance(node, ast.AsyncFunctionDef):
+                if node.col_offset == 0:
+                    func = Function(
+                        name=node.name,
+                        args=node.args,
+                        returns=TypeHint(annotation=node.returns) if node.returns else None,
+                        decorators=[],
+                        body=list(node.body),
+                        is_async=isinstance(node, ast.AsyncFunctionDef)
+                    )
+                    file_info.functions.append(func)
+
+        for node in ast.walk(tree):
+            if isinstance(node, ast.ClassDef):
+                cls = Class(
+                    name=node.name,
+                    bases=[ast.unparse(base) for base in node.bases],
+                    methods=[],
+                    attributes=[],
+                    decorators=[],
+                    body=list(node.body)
+                )
+
+                for item in node.body:
+                    if isinstance(item, ast.FunctionDef) or isinstance(item, ast.AsyncFunctionDef):
+                        method = Function(
+                            name=item.name,
+                            args=item.args,
+                            returns=TypeHint(annotation=item.returns) if item.returns else None,
+                            decorators=[],
+                            body=list(item.body),
+                            is_async=isinstance(item, ast.AsyncFunctionDef)
+                        )
+                        cls.methods.append(method)
+                    elif isinstance(item, ast.AnnAssign) and isinstance(item.target, ast.Name):
+                        var = Variable(
+                            name=item.target.id,
+                            annotation=TypeHint(annotation=item.annotation) if item.annotation else None
+                        )
+                        cls.attributes.append(var)
+
+                file_info.classes.append(cls)
+
+        return file_info
+
+    def generate_file(self, input_path: Path, output_path: Optional[Path] = None) -> Path:
+        """Generate a stub file for a single Python file."""
+        from stubgen.parser import FileParser
+        
+        file_info = FileParser(input_path).parse()
+        stub_content = self.writer.generate(file_info)
+        
+        if output_path is None:
+            output_path = input_path.with_suffix(".pyi")
+        
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+        
+        with open(output_path, 'w', encoding='utf-8') as f:
+            f.write(stub_content)
+        
+        return output_path
+    
+    def generate_directory(
+        self,
+        input_dir: Path,
+        output_dir: Optional[Path] = None,
+        recursive: bool = True
+    ) -> List[Path]:
+        """Generate stub files for all Python files in a directory."""
+        from stubgen.parser import parse_directory
+        
+        if output_dir is None:
+            output_dir = input_dir
+        
+        results = []
+        
+        file_infos = parse_directory(input_dir, recursive)
+        
+        for py_file, file_info in sorted(file_infos.items()):
+            try:
+                stub_content = self.writer.generate(file_info)
+                rel_path = py_file.relative_to(input_dir)
+                output_path = output_dir / rel_path.with_suffix(".pyi")
+                
+                output_path.parent.mkdir(parents=True, exist_ok=True)
+                
+                with open(output_path, 'w', encoding='utf-8') as f:
+                    f.write(stub_content)
+                
+                results.append(output_path)
+            except Exception as e:
+                print(f"Error generating stub for {py_file}: {e}", file=sys.stderr)
+        
+        return results