AI engine names
Wherever RSS Retriever asks for an AI engine (the Default AI model, AI Translate, AI Spinner, the engine parameter of [ai_generate] and [gpt_article]), you enter one string that names both the provider and the model. The plugin reads it and calls the right API with the right key.
The three formats
| Format | Example | Provider and key used |
|---|---|---|
openai-model | openai-gpt-5.4-mini | OpenAI, OpenAI key |
anthropic-model | anthropic-claude-sonnet-4-6 | Anthropic, Anthropic Claude key |
google-model | google-gemini-3.5-flash | Google Gemini, Google Gemini key |
xai-model | xai-grok-4.3 | xAI, xAI key |
endpoint ID-model | deepseek-deepseek-chat | Your custom endpoint with the ID deepseek |
| anything else | mistralai/mistral-large | OpenRouter, OpenRouter key; the string is the OpenRouter model ID |
The part after the prefix is the model name exactly as the provider spells it in its API documentation. New models work as soon as the provider releases them; you do not have to wait for a plugin update.
The built-in prefixes are checked first, then your endpoint IDs, and everything that matches neither is sent to OpenRouter. So do not name a custom endpoint openai, anthropic, google or xai.
Quick Select presets
The Quick Select button next to every AI engine field offers ready-made names. Each popular model is available twice: directly from its provider and through OpenRouter.
| Preset | Engine name |
|---|---|
| OpenAI GPT-5.5 | openai-gpt-5.5 / openai/gpt-5.5 |
| OpenAI GPT-5.4 mini | openai-gpt-5.4-mini / openai/gpt-5.4-mini |
| OpenAI GPT-4o mini | openai-gpt-4o-mini / openai/gpt-4o-mini |
| OpenAI GPT-4o | openai-gpt-4o / openai/gpt-4o |
| Anthropic Claude 4.7 Opus | anthropic-claude-opus-4-7 / ~anthropic/claude-opus-latest |
| Anthropic Claude 4.6 Sonnet | anthropic-claude-sonnet-4-6 / ~anthropic/claude-sonnet-latest |
| Anthropic Claude 4.5 Haiku | anthropic-claude-haiku-4-5 / ~anthropic/claude-haiku-latest |
| Google Gemini 3.1 Pro | google-gemini-3.1-pro-preview / ~google/gemini-pro-latest |
| Google Gemini 3.5 Flash | google-gemini-3.5-flash / ~google/gemini-flash-latest |
| xAI Grok | xai-grok-4.3 / x-ai/grok-4.3 |
| DeepSeek (OpenRouter) | deepseek/deepseek-v4-pro |
OpenRouter IDs that start with ~, such as ~anthropic/claude-sonnet-latest, always point to the latest model of that family.
Short aliases
For compatibility with older settings, a few short names after a provider prefix are translated into current models:
| You write | Model used |
|---|---|
anthropic-claude | claude-sonnet-4-5 |
anthropic-haiku | claude-haiku-4-5 |
openai-gpt4o | gpt-4o |
google-gemini-pro | gemini-3-pro-preview |
google-gemini-flash | gemini-3-flash-preview |
xai-grok | grok-4-1-fast-non-reasoning |
Choosing a model
- Translation, keywords, titles, image prompts: small, fast models such as GPT-4o mini, GPT-5.4 mini, Claude Haiku or Gemini Flash. They cost a fraction of the large models.
- Rewriting and long articles: larger models such as Claude Sonnet or Opus, GPT-5.5 or Gemini Pro write noticeably better text.
- Many models, one bill: OpenRouter gives access to models of all providers with one key and one balance.