Advanced Features
Override LLM Router
Override Storytell`s intelligent LLM router to specifically select which one to use in your responses.
Overriding the LLM Router
Storytell automatically selects the best AI model for your prompt using an intelligent LLM router. However, you can manually choose a specific model by including its keyword in your prompt.
Available LLMs
How to Override the LLM Router
To manually select a model, add its keyword anywhere in your prompt.
1
Open the prompt input bar
Navigate to your desired Collection or SmartChat™.
2
Craft your prompt
Type your request as usual.
3
Include the LLM keyword
Add the model’s keyword anywhere in the prompt.
4
Send your prompt
Press Enter or click the send button.
Keywords work anywhere in the prompt, regardless of their position.
Examples
Keyword Matching & Variations
You don’t need to type the exact name of a model. Storytell recognizes common variations automatically.
For example:
- “GPT-4” → Recognized as OpenAI’s GPT-4
- “Claude 3 Opus” → Recognized as Anthropic’s Claude 3 Opus
- “Llama 3 70B” → Recognized as Meta’s Llama 3
AI Model | Example Keywords | Accepted Variations |
---|---|---|
OpenAI (GPT) | gpt-4 , chatgpt , openai | gpt 4 , gpt-4o , chat gpt |
Google Gemini | gemini pro , gemini ultra , gemini | gemini pro , gemini-ultra , gemini |
Anthropic Claude | claude , claude 3 opus , sonnet | claude 3 , claude opus , haiku |
Meta (Llama) | llama 3 , llama 3 70b , meta | llama-3 , llama 3.2 7B , llama 3 70b |
DeepSeek | deepseek , deepseek r1 | deepseek-r1 |
Using multiple conflicting keywords may result in unexpected behavior. Use only one per prompt.
Visual Feedback
When keyword is detected, Storytell highlights it in your prompt to confirm recognition.