Translation Engines

The Cheat UI supports multiple translation engines ranging from free public APIs to high-performance local clusters and legacy Korean translators.

For instructions on how to actually start these servers, see the Server Setup (Run) Guide.

Comparison of Engines

Engine	Best For	Languages	Setup
Lingva	General Use	Most pairs (JA → EN focus)	Docker (Free)
ezTrans	Japanese to Korean	JA → KR ONLY	Local App (Windows)
LLMs	Dialogue / Immersion	All	API Key or Ollama

---

ezTrans (Japanese → Korean)

Legacy support for the popular Korean translation tools. This remains the fastest and most accurate choice for Korean speakers.

ezTransWeb

• Repo: HelloKS/ezTransWeb
• Setup:
  1. Install ezTrans XP on your Windows machine.
  2. Run the ezTransWeb python server.
  3. In Cheat UI, select the ezTransWeb (JP → KR) endpoint.
• Default URL: http://localhost:5000/translate

eztrans-server

• Repo: nanikit/eztrans-server
• Setup:
  • High-performance Rust-based alternative to ezTransWeb.
• Default URL: http://localhost:8000

---

Lingva Translate (Private & Free)

Lingva is an open-source front-end for Google Translate that doesn't track you and doesn't require an API key.

Public Nodes

By default, the plugin uses lingva.ml. While free, public nodes may rate-limit you during bulk translations.

Local Docker Cluster (Recommended)

For unlimited, high-speed translation, we recommend running Lingva locally:

1. Install Docker.
2. Run our pre-configured suite: docker-compose up -d.
3. In Cheat UI, select Local Lingva Docker.

---

LLM Engines (AI Translation)

For the highest quality translation that understands context and tone.

Cloud Providers

• OpenAI: Supports GPT-4o-mini (very cheap and fast).
• Google Gemini: Supports Gemini 2.0 Flash (free tier available).

Local LLM (Ollama)

Run your own AI locally with Ollama:

1. Install Ollama and pull a model (e.g., ollama pull personal/llama3-8b-instruct).
2. Point the Cheat UI to http://localhost:11434/v1.
3. Select the Ollama (Local) preset.