AI Services
Doco Translate supports a wide range of AI and large language model (LLM) services for high-quality, context-aware translations. These services require an API key and typically produce better results than basic translation engines.
Supported AI Services
| Service | Provider | API Key | Strengths |
|---|---|---|---|
| Gemini | Required | Multimodal capabilities, fast processing, generous free tier | |
| OpenAI | OpenAI | Required | High-quality translations, excellent context understanding |
| Claude | Anthropic | Required | Excellent for nuanced, context-aware translations |
| DeepSeek | DeepSeek | Required | Cost-effective, good performance for technical content |
| Zhipu AI | Zhipu AI | Required | Strong Chinese language support, free models available |
| Kimi | Moonshot AI | Required | Long context support, good for lengthy documents |
| MiniMax | MiniMax | Required | High-speed model variants available |
| Xiaomi MiMo | Xiaomi | Required | Good general-purpose performance |
| SiliconFlow | SiliconFlow | Required | Multi-model platform with free tiers (Qwen, DeepSeek, GLM) |
| OpenRouter | OpenRouter | Required | Unified interface to access models from multiple providers |
Click the service name to visit the provider's website and get an API key. You can also find these links in Settings → Services within the app via the Apply for API Key button.
Configuring an AI Service
Step 1: Get an API Key
Visit the provider's website using the links above to create an account and generate an API key.
Step 2: Enter the API Key in Doco Translate
- Go to Settings → Services.
- Select the AI service from the list on the left.
- Paste your API key in the API Key field.
- The key icon next to the service name changes from empty to filled, indicating the key is configured.
Step 3: Select a Model
Each AI service offers multiple models with different capabilities:
- In the service settings, find the Model dropdown.
- Select the model you want to use.
- Models marked as (Free) can be used without charges from the provider.
Tip: Smaller/faster models (e.g., flash, mini, haiku) are cheaper and faster but may produce slightly lower quality. Larger models (e.g., pro, opus) produce the best results but cost more.
Step 4: Verify
Click Verify service to test the connection with a sample translation.
Managing Models
Fetching the Model List
Some services support fetching the latest available models directly from the provider:
- Click Fetch model list in the service settings.
- The app connects to the provider and retrieves available models.
- New models not already in your list are presented for selection.
- Select the models you want to add and click Add.
Adding Custom Models
You can manually add models that aren't in the preset list:
- Click Add Model in the service settings.
- Enter the exact model identifier (e.g.,
gpt-4o-2024-08-06). - Optionally mark it as a free model.
- Click Add.
Deleting Custom Models
To remove a manually added model:
- Select the model in the model dropdown.
- Click Delete next to the model selector.
- Confirm the deletion.
Note: You can only delete custom (manually added) models. Preset models that ship with Doco Translate cannot be deleted.
Custom API Host
For advanced users, you can customize the API host for any service:
- In the service settings, find the API Host field.
- Enter the custom endpoint URL.
- The preview below shows the full request URL (host + path).
This is useful for:
- Using API proxies or mirrors
- Enterprise deployments with custom endpoints
- Self-hosted API-compatible services
Click the gear icon next to the API Host field to access advanced settings:
- API Path — The request path (e.g.,
/v1/chat/completions) - Request Body — Additional JSON fields merged into each request
- Request Headers — Additional HTTP headers
Restore Defaults
If you've customized a service's host, path, or other settings and want to revert:
- Click Restore Default at the bottom of the service settings.
- The host, path, request body, and headers are reset to the built-in defaults.
Your API key and model selection are preserved.
Previous: Built-in Services · Next: Local LLM Services
