UnityGameTranslator Files

v0.9.57

New Features

• Google Translate & DeepL API backends — Choose between LLM (AI), Google Translate, or DeepL for auto-translation. Configurable in Settings > Translation tab via a type selector (LLM / Translation API) and provider dropdown (Google / DeepL)
• Rate limit handling — HTTP 429 responses from any backend (LLM, Google, DeepL) are detected and failed translations are automatically re-queued with configurable retry delay (default 3s, supports decimals)
• Notification overlay settings — Disable the corner notification overlay entirely or choose its position (top-right, top-left, bottom-right, bottom-left) in Settings > General

Improvements

• Smart endpoint resolution — Automatically detects URL format for all OpenAI-compatible providers. Just paste the base URL from your provider's docs: | Provider | URL to enter | |----------|-------------| | Ollama | http://localhost:11434 | | LM Studio | http://localhost:1234/v1 | | Groq | https://api.groq.com/openai/v1 | | OpenRouter | https://openrouter.ai/api/v1 | | Together AI | https://api.together.xyz/v1 | | Gemini | https://generativelanguage.googleapis.com/v1beta/openai/chat/completions |

The mod auto-detects the URL format and resolves the correct endpoint. URLs ending with /v1 get /chat/completions appended. Bare URLs (like Ollama) get /v1/chat/completions. URLs with /v1/ followed by a wrong path are corrected. Full endpoint URLs are used as-is. Existing configurations are fully backwards compatible.

Note: Gemini uses a non-standard path (/v1beta/openai/chat/completions instead of /v1/chat/completions). Use the full URL above for Gemini.

Fixes

• Groq compatibility — Fix "property 'think' is unsupported" error by broadening think parameter detection for all provider error formats
• Login from wizard — Fix Connect Account button not working during wizard setup