Use a different model
Each model the agent can use is an LLM record in the Setup Portal. An agent uses the one named in its LLM field, so changing model is two steps: create the LLM record if it does not exist, then pick it on the agent.
Setup creates two to start with, Gemini 3.7 Flash and a free model router, both on OpenRouter.
Add a model
Section titled “Add a model”In the Setup Portal, open LLMs and select New.
| Field | What to enter |
|---|---|
| Model API Name | The model’s name exactly as the provider gives it. On OpenRouter, for example, google/gemini-3.7-flash |
| Named Credential | The connection to the provider. For the preconfigured one, OpenRouter |
| Chat Completions Path | The provider’s path. For OpenRouter, /api/v1/chat/completions |
| Family | The model’s maker, such as OpenAI, Anthropic or Google |
| Supports Reasoning? | Tick if the model can reason before answering |
| Token costs | The provider’s prices per million tokens for input, cached input and output. Tokens are the units models count text in; 1,000 is roughly 750 words. These drive the cost figures in the logs and dashboard |
A new model on the provider you already use needs nothing else. It reuses the key you added.
Switch an agent to it
Section titled “Switch an agent to it”Open the Agent record and set LLM to the new record. The next message uses it.
Two settings on the Agent record depend on the model, so check them when you switch:
- Reasoning Effort: set it to disabled for a model that does not support reasoning.
- Max Tokens: the longest reply the model may write.
See Customise an agent for what each does.
A model from another provider
Section titled “A model from another provider”To use a model from OpenAI, Azure OpenAI, your company’s AI gateway or anywhere else, connect that provider first. See Use a different provider.