Back to Nodes

Openrouter LLM

Last updated May 19, 2026

n8n community node for sending chat completion requests through OpenRouter.

49 Weekly Downloads
835 Monthly Downloads

Included Nodes

Openrouter LLM

Description

n8n-nodes-openrouter-custom-v2

n8n community package that adds an Openrouter LLM node for chat completions via OpenRouter — one HTTP API to route prompts to many hosted models.

Contents

  • Quick start
  • Credentials
  • Capabilities
  • Links
  • Quick start

    1. Install as a community node from the n8n UI, or add the package n8n-nodes-openrouter-custom-v2 to your environment (same guide).
    2. Create OpenRouter Custom V2 API credentials in n8n (Credentials).
    3. Add the Openrouter LLM node, attach those credentials, and pick a model (searchable list or model ID / expression).
    4. Set prompts with expressions so each input item drives its own request — one completion per item.

    For model IDs, pricing, and provider behaviour, see OpenRouter documentation.

    Credentials

    In n8n, create OpenRouter Custom V2 API credentials (tested against GET /models).

    | Field | Purpose |
    | ——– | ———————————————————————– |
    | API key | Bearer token for OpenRouter |
    | Base URL | OpenRouter-compatible API base (default https://openrouter.ai/api/v1) |
    | Site URL | Optional; sent for OpenRouter attribution (HTTP-Referer) |
    | App name | Optional; sent as X-OpenRouter-Title |

    Capabilities

    Chat and prompts

  • POST /chat/completions — Assistant text plus the raw API response on each output item.
  • Prompt modes — Simple prompt, system + user split, or full Messages JSON for multi-turn chats.
  • Generation — Sampling parameters and optional reasoning controls where the model supports them.
  • Models and routing

  • Model selection — Searchable list (from OpenRouter) or manual model ID / expression.
  • Model options — Fallback models and OpenRouter variants (e.g. :nitro, :online); conflicting combinations are validated.
  • Provider routing — Allow/deny providers, sort order, and related OpenRouter routing fields.
  • Structured output

  • JSON Schema — Optional schema with AJV validation and retries for machine-readable output.
  • Integrations and observability

  • Headers and metadata — Custom headers and request metadata where needed.
  • Tracing and search — Langfuse trace ID and web-search-related options when applicable.
  • Links

  • Installing community nodes (n8n)
  • OpenRouter documentation
  • CHANGELOG.md — release notes; see npm or this file for the current published version.
  • CONTRIBUTING.md — local development, compatibility, and changelog policy for contributors.

Package: n8n-nodes-openrouter-custom-v2 · License: MIT · Repository: SanderSpiegelaar/n8n-nodes-openrouter-custom-v2 · Contributing: CONTRIBUTING.md