Getting started with Wenyi
Wenyi translates books and long-form writing using your chosen model service. Start with the interface that fits your work; each interface keeps its own project state rather than sharing a live project database.
Choose an interface
| Interface | Start here | Best suited to |
|---|---|---|
| Desktop | Install and set up | Local projects, native file selection and an integrated proofreading workspace |
| Web | Deploy a browser workspace | A self-hosted service backed by PostgreSQL and Redis |
| CLI | Install and translate | Terminal workflows and automation without a server |
Desktop installers include the translation engine and Python runtime. You do not need the source-development environment to use a release.
Translate and revise a book
For Desktop and Web, the workspace guide follows a project from import and model selection through translation, proofreading, terminology, Review and export. Platform-specific installation, credentials and backups stay in the Desktop and Web guides.
Read the translation pipeline to understand preparation, context, three-draft translation, Review and resumable progress. These choices affect both quality and cost; model output still needs human review.
Look up a setting or format
- Configuration reference: providers, model routes, pipeline options and defaults.
- Format reference: supported inputs and outputs, PDF backends, layout limitations and bilingual exports.
Translation sends book text to your configured model provider unless you use a local model. PDF parsing can also use an external service. Check the provider's privacy terms and your rights to the text before sending a book.
Developing or packaging Wenyi? Use the separate development documentation.