diff --git a/docs/Makefile b/docs/Makefile
new file mode 100644
index 0000000..b34416c
--- /dev/null
+++ b/docs/Makefile
@@ -0,0 +1,30 @@
+# Makefile for Sphinx documentation
+
+# You can set these variables from the command line.
+SPHINXOPTS    =
+SPHINXBUILD   = sphinx-build
+SOURCEDIR     = .
+BUILDDIR      = _build
+
+# Put it first so that "make" without argument is like "make help".
+help:
+	@$(SPHINXBUILD) -M help "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
+
+.PHONY: help Makefile clean
+
+# Clean build directory
+clean:
+	rm -rf $(BUILDDIR)/*
+
+# Build HTML documentation
+html:
+	@$(SPHINXBUILD) -M html "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
+
+# Build and open HTML documentation
+html-open: html
+	open $(BUILDDIR)/html/index.html || xdg-open $(BUILDDIR)/html/index.html
+
+# Catch-all target: route all unknown targets to Sphinx using the new
+# "make mode" option.  $(O) is meant as a shortcut for $(SPHINXOPTS).
+%: Makefile
+	@$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
diff --git a/docs/conf.py b/docs/conf.py
new file mode 100644
index 0000000..8a62c13
--- /dev/null
+++ b/docs/conf.py
@@ -0,0 +1,128 @@
+"""
+Sphinx configuration file for Wan2.1 documentation.
+
+Copyright (c) 2025 Kuaishou. All rights reserved.
+"""
+
+import os
+import sys
+from datetime import datetime
+
+# Add source directory to path
+sys.path.insert(0, os.path.abspath('..'))
+
+# -- Project information -----------------------------------------------------
+
+project = 'Wan2.1'
+copyright = f'{datetime.now().year}, Kuaishou'
+author = 'Kuaishou'
+release = '2.1.0'
+version = '2.1'
+
+# -- General configuration ---------------------------------------------------
+
+extensions = [
+    'sphinx.ext.autodoc',
+    'sphinx.ext.napoleon',
+    'sphinx.ext.viewcode',
+    'sphinx.ext.intersphinx',
+    'sphinx.ext.mathjax',
+    'sphinx.ext.todo',
+    'sphinx.ext.coverage',
+    'sphinx.ext.githubpages',
+    'myst_parser',  # For markdown support
+]
+
+# Add any paths that contain templates here, relative to this directory.
+templates_path = ['_templates']
+
+# List of patterns, relative to source directory, that match files and
+# directories to ignore when looking for source files.
+exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
+
+# The suffix(es) of source filenames.
+source_suffix = {
+    '.rst': 'restructuredtext',
+    '.md': 'markdown',
+}
+
+# The master toctree document.
+master_doc = 'index'
+
+# -- Options for HTML output -------------------------------------------------
+
+html_theme = 'sphinx_rtd_theme'
+html_static_path = ['_static']
+html_logo = None
+html_favicon = None
+
+html_theme_options = {
+    'canonical_url': '',
+    'analytics_id': '',
+    'logo_only': False,
+    'display_version': True,
+    'prev_next_buttons_location': 'bottom',
+    'style_external_links': False,
+    'style_nav_header_background': '#2980B9',
+    # Toc options
+    'collapse_navigation': True,
+    'sticky_navigation': True,
+    'navigation_depth': 4,
+    'includehidden': True,
+    'titles_only': False
+}
+
+# -- Extension configuration -------------------------------------------------
+
+# Napoleon settings (for Google/NumPy style docstrings)
+napoleon_google_docstring = True
+napoleon_numpy_docstring = True
+napoleon_include_init_with_doc = True
+napoleon_include_private_with_doc = False
+napoleon_include_special_with_doc = True
+napoleon_use_admonition_for_examples = False
+napoleon_use_admonition_for_notes = False
+napoleon_use_admonition_for_references = False
+napoleon_use_ivar = False
+napoleon_use_param = True
+napoleon_use_rtype = True
+napoleon_preprocess_types = False
+napoleon_type_aliases = None
+napoleon_attr_annotations = True
+
+# Autodoc settings
+autodoc_default_options = {
+    'members': True,
+    'member-order': 'bysource',
+    'special-members': '__init__',
+    'undoc-members': True,
+    'exclude-members': '__weakref__'
+}
+autodoc_typehints = 'description'
+autodoc_typehints_description_target = 'documented'
+
+# Intersphinx mapping
+intersphinx_mapping = {
+    'python': ('https://docs.python.org/3', None),
+    'torch': ('https://pytorch.org/docs/stable/', None),
+    'numpy': ('https://numpy.org/doc/stable/', None),
+}
+
+# Todo extension
+todo_include_todos = True
+
+# MyST parser options
+myst_enable_extensions = [
+    "colon_fence",
+    "deflist",
+    "dollarmath",
+    "fieldlist",
+    "html_admonition",
+    "html_image",
+    "linkify",
+    "replacements",
+    "smartquotes",
+    "strikethrough",
+    "substitution",
+    "tasklist",
+]
diff --git a/docs/index.rst b/docs/index.rst
new file mode 100644
index 0000000..e6406e7
--- /dev/null
+++ b/docs/index.rst
@@ -0,0 +1,157 @@
+Wan2.1 Documentation
+====================
+
+Welcome to the Wan2.1 documentation! Wan2.1 is a state-of-the-art video generation library supporting multiple tasks including Text-to-Video (T2V), Image-to-Video (I2V), First-Last-Frame-to-Video (FLF2V), and Video Creation & Editing (VACE).
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Getting Started
+
+   installation
+   quickstart
+   tutorials/index
+
+.. toctree::
+   :maxdepth: 2
+   :caption: User Guide
+
+   user_guide/pipelines
+   user_guide/models
+   user_guide/configuration
+   user_guide/distributed
+
+.. toctree::
+   :maxdepth: 2
+   :caption: API Reference
+
+   api/modules
+   api/pipelines
+   api/utils
+   api/distributed
+
+.. toctree::
+   :maxdepth: 1
+   :caption: Development
+
+   contributing
+   changelog
+   license
+
+Indices and tables
+==================
+
+* :ref:`genindex`
+* :ref:`modindex`
+* :ref:`search`
+
+Quick Links
+===========
+
+- `GitHub Repository <https://github.com/Kuaishou/Wan2.1>`_
+- `Issue Tracker <https://github.com/Kuaishou/Wan2.1/issues>`_
+- `PyPI Package <https://pypi.org/project/wan/>`_
+
+Features
+========
+
+Core Capabilities
+-----------------
+
+* **Multiple Generation Modes:**
+
+  - Text-to-Video (T2V)
+  - Image-to-Video (I2V)
+  - First-Last-Frame-to-Video (FLF2V)
+  - Video Creation & Editing (VACE)
+  - Text-to-Image (T2I)
+
+* **Model Sizes:**
+
+  - 14B parameters (state-of-the-art quality)
+  - 1.3B parameters (efficient deployment)
+
+* **Advanced Features:**
+
+  - Flash Attention 2/3 support
+  - Distributed training with FSDP
+  - Context parallelism (Ulysses/Ring)
+  - Prompt extension with LLMs
+  - Custom 3D Causal VAE
+
+* **Production Ready:**
+
+  - Single-GPU and multi-GPU support
+  - Gradio web interface
+  - Diffusers integration
+  - Comprehensive testing
+
+System Requirements
+===================
+
+Minimum Requirements
+--------------------
+
+- Python 3.10+
+- PyTorch 2.4.0+
+- CUDA 11.8+ (for GPU support)
+- 24GB+ GPU memory (for 1.3B model)
+- 80GB+ GPU memory (for 14B model)
+
+Recommended
+-----------
+
+- Python 3.11
+- PyTorch 2.4.1
+- CUDA 12.1
+- NVIDIA A100 (80GB) or H100
+
+Quick Start
+===========
+
+Installation
+------------
+
+.. code-block:: bash
+
+   pip install wan
+
+Basic Usage
+-----------
+
+.. code-block:: python
+
+   from wan.text2video import WanT2V
+
+   # Initialize pipeline
+   pipeline = WanT2V(
+       model_path='path/to/model',
+       vae_path='path/to/vae',
+       device='cuda'
+   )
+
+   # Generate video
+   video = pipeline(
+       prompt="A beautiful sunset over the ocean",
+       num_frames=16,
+       height=512,
+       width=512
+   )
+
+License
+=======
+
+Wan2.1 is released under the Apache 2.0 License. See the LICENSE file for details.
+
+Citation
+========
+
+If you use Wan2.1 in your research, please cite:
+
+.. code-block:: bibtex
+
+   @software{wan2024,
+     title={Wan2.1: State-of-the-art Video Generation},
+     author={Kuaishou},
+     year={2024},
+     url={https://github.com/Kuaishou/Wan2.1}
+   }