WeClone is an open-source, one-stop solution for creating an AI twin of yourself from your own chat history. With more than 18,300 stars on GitHub, it covers the full lifecycle of a digital avatar: exporting and cleaning chat logs, fine-tuning a large language model on your messages so it absorbs your speaking style, and deploying the result as a chatbot that talks like you. The project is actively maintained, with updates landing in October 2026.
WeClone starts from the conversations you already have โ WeChat, Telegram, and other chat exports โ and preprocesses them into training data with privacy filtering built in. It then fine-tunes open models such as Qwen on that data, capturing your vocabulary, tone, and conversational habits rather than generic assistant-speak. The trained avatar can be connected to Telegram, Discord, and Slack for multi-platform interaction, and the whole pipeline supports local deployment and local fine-tuning, so personal data stays under your control instead of being uploaded to a third-party clone service.
WeClone is free and open source under the AGPL-3.0 license. Your only costs are the computing resources you choose to train and host the avatar on.
WeClone is for hobbyists, creators, and digital-legacy enthusiasts who want a personal AI double trained on their own words, and for developers exploring style-personalized fine-tuning. Because training and deployment can stay entirely on your own machines, it is also a fit for anyone curious about digital avatars who does not want to upload years of private conversations to a commercial service.