When configuring an Agent, the underlying Large Language Model (LLM) that powers its responses may be important depending on your needs.
We provide you with two options whether you want to use Agents within le Chat or via our APIs.
Agents created in le Chat
Agents created directly within the le Chat interface are focus on providing a seamless, easy-to-setup and intuitive conversational experience.
As such, there is currently no option to select a specific or custom model for Agents built in le Chat.
🔑 These Agents are automatically powered by the same latest, high-performance Large Language Models (LLMs) that drive the general le Chat conversational experience.
Agents created via la Plateforme Agent Builder (for API use)
On the other hand, Agents created using the Agent Builder on la Plateforme, typically for deployment and use via the API, may need further customization and control over the underlying model.
On la Plateforme, you may select a specific Mistral model, or even one of your fine-tuned models to power your Agent.
🔑 This allows you to optimize performance, cost, and response quality based on your specific use case.
🔎 For detailed information on creating Agents for API use and interacting with them programmatically, please refer to the official API documentation.