n8n-nodes-w1-gateway

n8n community node – connects AI agents and AI nodes in n8n to the W1 LiteLLM Gateway so your workflows can use any model exposed by the gateway.

Overview

The W1 Gateway Chat Model node provides a Language Model provider for n8n's AI/Agent nodes.
It communicates with the W1-hosted LiteLLM service, which aggregates dozens of LLM providers behind a single OpenAI-compatible endpoint.

The gateway URL is hardcoded – you only need to supply your API token.

Installation

Via n8n UI (recommended)

1. Open your n8n instance → Settings → Community Nodes → Install.
2. Enter n8n-nodes-w1-gateway and click Install.

Manual (npm)

cd ~/.n8n
npm install n8n-nodes-w1-gateway

Credentials

Create a new W1 Gateway API credential and enter your API Key (token).

No URL configuration is needed – the gateway endpoint is fixed.

Node: W1 Gateway Chat Model

The node outputs an AI Language Model that can be connected to:

• AI Agent node
• Basic LLM Chain node
• Any other n8n LangChain-compatible chain/agent node

Parameters

| Parameter | Description | Default | |---|---|---| | Model | Model name available in the gateway (e.g. gpt-4o-mini, claude-3-5-sonnet) | gpt-4o-mini | | Temperature (Options) | Sampling temperature (0–2) | 0.7 | | Maximum Tokens (Options) | Max tokens to generate (-1 = model default) | -1 | | Top P (Options) | Nucleus sampling probability mass | 1 | | Frequency Penalty (Options) | Penalise repeated tokens (-2–2) | 0 | | Presence Penalty (Options) | Penalise tokens already present (-2–2) | 0 | | Max Retries (Options) | Number of retry attempts on failure | 2 | | Timeout (ms) (Options) | Request timeout in milliseconds | 60000 |

Example workflow

┌──────────────────────┐        ┌────────────────────────────┐
│  W1 Gateway          │──────▶ │  AI Agent / LLM Chain      │
│  Chat Model          │ Model  │                            │
└──────────────────────┘        └────────────────────────────┘

License

MIT – © W1 AI (w1-ai.com)