[New Community Node] The Ultimate OpenRouter Toolkit: Unified Multimodal, Embeddings & Reranker in one API key!

Built with n8n
, , ,
1

Hi fellow automators! :waving_hand:

I’m Jay Nguyen (Verified n8n Creator & Founder of GenStaff), and today I’m thrilled to share a massive upgrade for your AI Agent workflows: n8n-nodes-openrouter-official.

If you’ve been building Advanced AI (LangChain) workflows in n8n, you’ve probably hit these frustrating bottlenecks:

  1. Credential Fragmentation: Setting up separate billing, API keys, and native nodes for OpenAI, Anthropic, Google, and open-source models is a nightmare.
  2. Multimodal Limitations: Feeding complex documents (like PDFs, Video, or Audio) into an LLM often requires clunky workarounds.
  3. The Reranker Void: High-quality RAG (Retrieval-Augmented Generation) requires Reranking to fight hallucinations, but n8n natively lacks accessible Document Compressors/Rerankers outside of Cohere.

I built this Ultimate OpenRouter Toolkit to solve all of the above.

With just one OpenRouter API Key, this package unlocks three incredibly powerful nodes that seamlessly integrate into your n8n canvas:

:glowing_star: 1. The Unified OpenRouter Action Node

Stop jumping between different provider nodes. This single node gives you access to hundreds of models instantly.

  • True Multimodal Power: We’ve designed this node to easily digest PDF Documents, Images, Audio, and Video. Just map your binary data, and as long as the model (like Gemini 3.5 Flash or Claude 5 Sonnet…) supports it, the node handles the rest.
  • It also natively supports routing requests for Image Generation and Video Generation right on the canvas.

:brain: 2. OpenRouter Embeddings (Advanced AI)

Break free from the OpenAI embedding lock-in. This node plugs directly into n8n’s Vector Store ecosystem, allowing you to generate embeddings using any of the highly efficient, cost-effective models available on OpenRouter.

:bullseye: 3. OpenRouter Reranker (Document Compressor)

This is the game-changer for RAG. Retrieving documents often yields bloated, irrelevant context. With this node, you can now use OpenRouter’s most powerful reasoning LLMs as a Document Compressor. It will dynamically score, rerank, and filter your retrieved documents before they reach your final Agent—drastically reducing token costs and completely eliminating hallucinations.

:package: How to install

Head over to Settings > Community Nodes in your n8n instance and install:

n8n-nodes-openrouter-official

:link: NPM: n8n-nodes-openrouter-official :link: GitHub: nguyenthieutoan/n8n-nodes-openrouter-official

I can’t wait to see the next-level AI Agents you build with this. Drop a comment below if you have any questions, feature requests, or just want to share your workflow! Happy automating! :rocket:

image
image522×588 20.3 KB

image
image605×497 35 KB

image
image610×659 29.7 KB

2 Likes
2

Appreciate this! I’ve been using embeddings from openrouter through the HTTP Node so this will be much cleaner.

1 Like
3

@christopherjnelson You’re welcome! If you spot any HTTP request that should really be a node, just tell me and I’ll “nodemize” it for you.

1 Like
4

Also, I’ve built a few n8n community nodes out of my own production needs, so these are nodes that I actively use and that have passed n8n’s testing standards.
You can check them out here for reference: npm profile.
You can also just search for my name “nguyenthieutoan” (Nguyễn Thiệu Toàn) on npm or GitHub to see the full list of packages and projects I maintain.

image
image1710×1222 219 KB