Skip to content

API_GUIDE.md

Latest commit

 

History

History
52 lines (37 loc) · 2.23 KB

API_GUIDE.md

File metadata and controls

52 lines (37 loc) · 2.23 KB
Raw

API Configuration Guide

To use AI-powered translation providers, you need a valid API key from the respective provider.

🔑 Getting an API Key

Below are the most common providers and how to get their API keys.

Provider Access How to Get
Google Translate Free No API key needed
Microsoft Translator Free No API key needed
Yandex Translate Free No API key needed
Lingva Translate Free No API key needed
Google Gemini Free/Paid Google AI Studio
DeepL Free/Paid DeepL API Keys
OpenRouter Paid OpenRouter API Keys
OpenAI Paid OpenAI API Keys
DeepSeek Paid DeepSeek API Keys
WebAI to API Free Local Server

Multiple API Keys Support

You can enter multiple API keys for each provider (one per line). The extension will:

  • ✅ Automatically try the next key if the current one fails (e.g., quota limit).
  • ✅ Reorder working keys to the top for better performance.

🛠 Advanced Provider Settings

You can customize each provider's settings to balance quality and cost:

  • Google Gemini: Change the API URL to use different models (e.g., gemini-2.0-flash).
  • OpenAI: Enter specific model names like gpt-4o or gpt-3.5-turbo.
  • OpenRouter: Support for a variety of models; use the model name from the OpenRouter Docs.
  • DeepSeek: Customize model settings as per DeepSeek Docs.
  • DeepL: Configure API tier (Free or Pro) and formality settings.
  • Custom (OpenAI Compatible): Connect to self-hosted models (e.g., Ollama) or any third-party OpenAI-compatible API by providing a custom API URL and Model Name.

Default Models

If you don't specify a model, these are the defaults used:

  • OpenAI / OpenRouter: gpt-3.5-turbo
  • Google Gemini / WebAI to API: gemini-1.5-flash
  • DeepSeek: deepseek-chat