SuperAPI AI Directory
Partner with us中文AI newsFavorites
DirectoryNews and tutorials

Metasequoia IME - Free and open-source Chinese input method with AI association

Author: AI小集 · Original source: AI工具集

Metasequoia IMEWhat is it

Metasequoia IME is a free and open-source Chinese input method based on the Windows native TSF framework, designed specifically for Win10/Win11. Unlike most RIME-based solutions, the tool is completely self-developed and natively compatible with UWP and high-DPI multi-screen environments. Metasequoia IME supports full pinyin, double pinyin, and Wubi, and can enable voice input and AI association through a self-configured API. The core input logic runs locally, with no ads and no privacy theft, making it a new choice for Windows users seeking a pure and secure input experience.

Metasequoia IME

Metasequoia IMEMain features

  • Multi-scheme input: Supports three mainstream input schemes: full pinyin, double pinyin, and 86/98 Wubi.
  • Auxiliary code filtering: Supports auxiliary code schemes such as Lantian Xiaoyudian and Ziranma, accurately locating duplicate codes.
  • Super abbreviated pinyin and quick phrases: Supports custom quick phrases, improving the input efficiency of high-frequency expressions.
  • Voice input: Invokes recording via the right-click menu; you need to configure an ASR API yourself to implement speech-to-text.
  • AI-assisted association: Asynchronously generates a third intelligent candidate during full pinyin/double pinyin input, and supports custom prompts and LLM APIs such as DeepSeek.
  • Text polishing: Supports connecting to an LLM API for real-time polishing of input content.
  • Handwriting recognition pad: Provides a mouse/touch handwriting input panel that automatically recognizes candidate Chinese characters.
  • On-screen keyboard: Supports text and shortcut key input via mouse or touch.
  • Emoji and kaomoji panel: Built-in quick input of emojis.
  • Custom dictionaries: Supports custom import and addition of full pinyin dictionaries, Wubi dictionaries, and English dictionaries.
  • Appearance and shortcuts: Supports dark/light themes, customization of candidate window styles, and configuration of shortcuts such as paging/editing.
Wawa GitHub

Follow on WeChat and reply "open source" to join theAI open-source project communication group

How to use Metasequoia IME

  • Prepare the environment: Install Visual Studio 2026, CMake, vcpkg, Python 3.10+, and gsudo (installed via Scoop).
  • Clone the repository: Run git clone --recursive https://github.com/metasequoiaime/MetasequoiaImeTsf.git。
  • Initialize the environment: After entering the directory, run python .\scripts\prepare_env.py to automatically configure dependencies.
  • Compile the DLL: Respectively run .\scripts\lcompile.ps1 64 and .\scripts\lcompile.ps1 32 to generate dual-architecture components.
  • Deploy files: As administrator, create the C:\Program Files\MetasequoiaImeTsf and C:\Program Files (x86)\MetasequoiaImeTsf directories, and copy the DLLs for the corresponding bitness into each.
  • Register the input method: In an administrator PowerShell, run sudo regsvr32 .\MetasequoiaImeTsf.dll to complete system registration.
  • Configure APIs: In the settings panel, fill in the endpoint addresses and Tokens for the ASR API (voice) and LLM API (AI association/polishing) to enable advanced features.

Core advantages of Metasequoia IME

  • Native TSF architecture: Deeply adapted to modern Windows applications, with better compatibility than traditional IMM32 and cross-platform solutions.
  • Fully open-source and self-developed: Does not depend on the RIME engine, with independently controllable code, filling the gap of a purely self-developed open-source TSF input method for Windows.
  • Ultimate privacy and security: Core input logic runs locally, with no cloud dictionary uploads, no ad pop-ups, and no privacy theft.
  • Pluggable AI capabilities: Both voice and AI suggestions are integrated through open APIs, allowing users to choose their own service providers and models, offering flexibility and low cost.
  • Dual-architecture compatibility: Provides both 64-bit and 32-bit components, ensuring stable operation across various new and old software.

Project address of the Metasequoia Input Method

  • GitHub repository：https://github.com/metasequoiaime/MetasequoiaImeTsf

Application scenarios of the Metasequoia Input Method

  • Privacy-first office scenarios: Suitable for positions sensitive to input data security, such as government, finance, and R&D. Core input runs entirely locally, eliminating cloud uploads and privacy theft by commercial input methods.
  • Heavy users of the native Windows environment: For Win10/Win11 users who frequently use UWP apps, high-resolution screens, or multi-monitor office setups, the TSF architecture provides more stable compatibility and rendering experience than traditional input methods.
  • Advanced Shuangpin and Wubi input users: Supports auxiliary codes for precise filtering of duplicate codes and super abbreviated spelling, satisfying professional writers who pursue efficient shape-code input and need customizable shortcut phrases.
  • AI input method early adopters and customization: Through self-configured LLM APIs such as DeepSeek and ASR services, users can experience AI suggestions and voice input at low cost, suitable for tech enthusiasts who enjoy tinkering with tools.