AI Engines

AI Translation Engines

RenLocalizer supports modern Large Language Models (LLMs) to provide context-aware, high-quality translations.

🟢 OpenAI / OpenRouter

Standard for high-quality translations.

• Set up: Enter your API key in the settings.
• OpenRouter: You can use OpenRouter to access models like Claude 3.5 or Llama 3 via the OpenAI provider by changing the Base URL to https://openrouter.ai/api/v1.

🟡 Google Gemini

Fast and often has a generous free tier for developers.

• Provider: Select Gemini.
• API Key: Get one from Google AI Studio.
• Safety Settings: You can configure the safety threshold (BLOCK_NONE is recommended for uncensored game content).

🔵 Local LLM (Offline)

Run models locally on your own hardware for maximum privacy and zero cost.

• Backends: Supports Ollama, LM Studio, and LocalAI.
• Model Recommendation: Qwen 2.5 7B, Llama 3.1 8B, or Dolphin-Mistral (for uncensored content).
• Default URL: http://localhost:11434/v1 (Ollama) or http://localhost:1234/v1 (LM Studio).

⚙️ Advanced AI Parameters

Parameter	Default	Description
Temperature	0.3	Lower values (0.1-0.3) make translation consistent. Higher values make it creative.
Max Tokens	2048	Max length of each translation response.
Batch Size	5-10	Number of strings sent in a single prompt. Higher is faster but riskier for small models.
System Prompt	Auto	A custom persona instruction for the AI (e.g., "You are a professional localizer").

🛡️ Content Safety & Censorship

Unlike traditional engines, AI models might refuse to translate "sensitive" content.

• Gemini: Set Safety Level to BLOCK_NONE.
• Local LLMs: Use "Uncensored" or "Instruct" fine-tuned models to bypass ethical filters.
• Fallback: If an AI refuses a string, RenLocalizer can automatically fallback to Google Translate to ensure 100% completion.