Hello n8n Community,
I’ve been working with n8n’s AI Agent node and wanted to integrate a custom language model for an AI agent workflow. While n8n currently provides a range of built-in integrations for AI models, there seems to be no clear method to register or use custom AI nodes (such as custom language models or chat models) in the AI Agent node.
Problem:
When I create a custom node, for example, an integration with StraicoChatModel (or any other external AI), I am able to see it in the nodes list. However, I am unable to use it as a chat model in the AI Agent node or connect it directly to workflows that involve AI Agents.
Expected Behavior:
- I would expect to be able to add custom AI nodes to the AI Agent node, similar to how existing AI models like OpenAI or GPT models are integrated.
- The node should be recognized as a chat model and should be available in the AI Agent’s model selection dropdown.
- Custom nodes should have the ability to interact with the AI Agent just like built-in integrations, without needing to manually handle HTTP requests.
Request:
- Could we get official support for custom AI nodes to be integrated into the AI Agent node?
- If possible, a standard template or guideline on how to create and register custom chat models compatible with the AI Agent node would be extremely helpful.
I believe this would allow for much greater flexibility in workflows, enabling users to integrate their own language models and AI services with n8n’s AI Agent.
Thanks in advance for considering this feature request!