Memori Chat Model icon

Memori Chat Model

OpenAI-compatible chat model that injects memori_attribution (entity_id, process_id, session_id) into every request for self-hosted Memori

Memori Chat Model

AI-generated

Summary

Configure and use a Memori-hosted OpenAI-compatible chat model to generate chat completions with attribution metadata for entity, process, and session, optionally enabling reasoning and controlling privacy features.

Inputs

  • Model (required) — Select the Memori model by name from a dynamically loaded list or enter a custom model ID.
  • Entity ID (required) — Specify the Memori entity ID, typically the end-user, used for attribution (supports expressions).
  • Process ID (required) — Set the logical application or process name used for attribution, defaulting to 'n8n_agent'.
  • Session ID (required) — Provide the conversation or session identifier used for attribution (supports expressions).
  • Enable Thinking — Toggle to allow the model to emit reasoning tokens, recognized by certain vLLM/SGLang models.
  • Incognito — Set to bypass Memori's recall injection and database writes, disabling augmentation for the request.
  • Options — Configure advanced options including Base URL override, max retries, maximum tokens generated, sampling temperature, and HTTP timeout.

Output shape

a single AI language model instance

Returns a configured ChatOpenAI model instance from the Langchain library that includes memori_attribution metadata in requests. This node outputs no message texts itself but provides the configured model for downstream AI interactions in the workflow.

Links

Discussion