Google Gemini Chat Model icon

Google Gemini Chat Model

Chat Model Google Gemini

Google Gemini Chat Model

AI-generated

Summary

Provides a Google Gemini chat model to an n8n AI Agent or Chain. Choose a Gemini 2.5+ model and optionally configure generation, reasoning, output format, retry, and safety settings.

Inputs

  • Model (required) — Selects the Gemini model used to generate completions. The available model list is loaded from Google and filtered to exclude embedding and Imagen models and models older than Gemini 2.5. The default is Gemini 2.5 Flash for node version 1 and Gemini 3 Flash Preview for version 1.1 or later.
  • Options — Optional generation settings: maximum output tokens (default 1024), temperature (0.7), top K (40), top P (0.9), and up to five stop sequences. Configure thinking level for Gemini 3+ or thinking budget for Gemini 2.5, and choose whether to include thought summaries. JSON output can be enabled for supported models. Maximum retries defaults to 2 for transient errors. Safety settings accept one or more category and blocking-threshold pairs; the default is dangerous content blocked at the medium threshold and above.

Output shape

A single AI language model connection output for an Agent or Chain.

This is a model sub-node and must be connected to an AI Agent or Chain; it does not process standalone workflow items or return a completion directly.

Links

Discussion