TIPThe project is open-sourced on GitHub. You can visit Tokisaki-Galaxy/hexo-translate-llm for more information.
Cross-language information dissemination is increasingly important in the digital age. Manually translating articles is not only time-consuming but also hard to keep up with the update frequency. The Hexo LLM Translate plugin provides an automated solution for this. It uses large language models (LLMs) to deliver high-quality translation of blog content.
Core features
Smart bilingual switching improves the reading experience. The plugin detects the browser’s language preference, so readers can read the matching version without manually clicking anything.
The caching mechanism ensures translation efficiency. The plugin identifies changes via content hashes, and only triggers API calls when the body or title changes. This effectively reduces API usage costs.
Multi-device sync supports a remote database. Through Neon PostgreSQL, build caches can be shared in real time between different devices. (Optional)
Stability is at the core of the plugin’s design. A built-in rate limiter prevents the API from tripping under excessive concurrency. An automatic retry mechanism effectively handles network fluctuations.
SEO optimization ensures search rankings are preserved. The translated content is injected directly into the HTML structure, allowing search engines to fully crawl the bilingual information. Hexo-specific tags are also properly protected to avoid breaking the page structure.
For articles you don’t want translated, you can exclude them by setting no_translate: true in the Front-matter.
Quick start
Install the plugin
Run the install command in the Hexo root directory:
npm install hexo-translate-llmConfiguration
Add the following configuration to _config.yml. It’s recommended to choose a more cost-effective model:
llm_translation: enable: true model: deepseek-ai/DeepSeek-V3.2 endpoint: https://api.siliconflow.cn/v1/chat/completions max_concurrency: 4 single_timeout: 120NOTEIf you don’t want to use a remote database, you can skip the Neon PostgreSQL configuration step. The plugin will automatically use the local cache.
Security settings
This project manages sensitive information via environment variables.
Vercel deployment
In the project settings on the Vercel dashboard, add the LLM_API_KEY environment variable.
Set up a Storage connection to connect to the neon PostgreSQL database. Vercel will automatically inject the DATABASE_URL environment variable.
Local build
Create a .env file in the root directory:
LLM_API_KEY=您的密钥DATABASE_URL=数据库连接地址Usage tips
Specific articles can be excluded from translation. Just set no_translate: true in the Front-matter. The plugin automatically syncs the page’s title tag to keep the browser tab’s language consistent.