Immersive Translate with a translation API
Updated 2026-10-18 · Side-by-side translation of web pages and PDFs with a dedicated translation model: fast, pay per token, a few settings.
Connection details
- Base URL (OpenAI compatible)
- https://api.aifans.fan/v1
- API key
- Create one in the console; it starts with sk- (no account yet? sign up)
- Recommended models
- hy-mt2-pro, deepseek-v4.1-flash (all models and prices)
Immersive Translate is a widely used browser extension for side-by-side translation of web pages, PDFs and subtitles. It accepts a custom OpenAI-compatible API, so you can use AIFans API and pay per token.
Setup
- Open Immersive Translate's settings → Translation Services, find OpenAI (or a custom AI service) and open its settings.
- API key: your key.
- Model: choose a custom model and enter hy-mt2-pro; for long or tricky texts you can switch to deepseek-v4.1-flash.
- Custom API endpoint: https://api.aifans.fan/v1/chat/completions
- Set the maximum requests per second to 1–2, save, and make it your default service.
Why hy-mt2-pro
- A dedicated translation model: no “thinking”, fast (about 1 second per paragraph in our test), and it keeps the paragraph breaks.
- Low price; see its model page.
- Its context is short, which suits paragraph-by-paragraph translation; use a chat model for summaries or Q&A.
The extension's menus change between versions. Too many requests per second can cause 429 (too many requests); the extension retries, and throttled calls are not billed.