How Translation Works

Chapters translate as you click through, not all upfront. This keeps the queue moving and lets you start reading a long novel without waiting for the whole thing to finish first.

How It Works

When a novel is added to the site:

1. The scraper grabs every chapter (so the chapter list is complete from day one)
2. The first 50 chapters translate automatically in the background
3. Chapter titles for the rest of the novel batch-translate separately, so the chapter list is browsable end-to-end even before the bodies are translated

When you reach a chapter past the auto-translated window:

1. The translation runs the moment you open it (typically 5–10 seconds)
2. You'll see a streaming progress indicator while text fills in live, instead of a blank page
3. As soon as you start reading, the next 3 chapters prefetch in the background — by the time you click "Next", they're usually ready instantly

The first chapter you hit past chapter 50 is the only one with a noticeable wait. Everything after that should feel seamless thanks to the prefetch.

Tier Requirements

On-demand translation past the first 50 chapters requires an active subscription — at minimum the Reader tier ($4.99/mo). The first 50 chapters of every novel remain free for everyone.

When to Use Retranslate Instead

If a chapter is already translated but reads poorly, on-demand translation won't help — it only runs once per chapter. To rerun the AI on an existing translation (with the latest model and glossary), use the chapter's Retranslate button. See Retranslate & Rescrape for the walkthrough.

Why Demand-Based?

Most novels on the site have between 1,000 and 6,000 chapters, but most readers only ever reach the first 50–200. Translating every chapter upfront would burn LLM budget on chapters nobody reads. Demand-based translation means quota goes where the reading actually happens.