This is an n8n community node package. It lets you use GreenPT chat, embedding, reranking and speech-to-text models in your n8n workflows, AI agents and chains.
GreenPT is a sustainable AI platform that serves open-weight models through an OpenAI-compatible API, hosted in Europe, with a US region available.
n8n is a fair-code licensed workflow automation platform.
Installation Nodes Credentials Compatibility Usage Resources
Follow the installation guide in the n8n community nodes documentation and install @greenpt/n8n-nodes-greenpt.
| Node | Type | Use it for |
|---|---|---|
| GreenPT Chat Model | Chat model sub-node | The Chat Model input of an AI Agent, Basic LLM Chain or other AI root node |
| Embeddings GreenPT | Embeddings sub-node | The Embedding input of a vector store, to insert and to query documents |
| Reranker GreenPT | Reranker sub-node | The Reranker input of a vector store with Rerank Results turned on |
| GreenPT | Action node, also usable as an AI Agent tool | Create Embedding, Rerank Documents and Transcribe Audio |
Every model list loads live from the region set in your credential, so new GreenPT models appear without updating the package. The chat model list leaves out style variants (-caveman, -honey, -ponytail and their -lite/-ultra levels) and -raw models; select Expression and type the ID to use one.
| Parameter | Description |
|---|---|
| Model | A GreenPT chat model. Defaults to glm-5.3. |
| Maximum Number of Tokens | Upper limit on generated tokens. Reasoning models spend part of this budget on thinking. |
| Max Retries | Retries after a failed request. |
| Sampling Temperature | Randomness of the output, from 0 to 2. |
| Timeout | Request timeout in milliseconds. |
| Top P | Nucleus sampling. Change this or temperature, not both. |
Defaults to green-embedding. Use the same embedding model to insert documents and to query them.
Defaults to green-rerank. Top N sets how many documents the vector store returns after reranking; each keeps its relevanceScore in its metadata.
| Operation | Input | Output |
|---|---|---|
| Create Embedding | Text | embedding, model, usage |
| Rerank Documents | Query, a JSON array of document texts, Top N | One item per document: index, relevanceScore, document |
| Transcribe Audio | An audio file in a binary field, language code; options to identify speakers, punctuate and smart-format | transcript, confidence, duration, words |
- Create a GreenPT account at account.greenpt.ai. Every plan includes API access.
- Create an API key at account.greenpt.ai/api/keys.
- In n8n, create a GreenPT API credential and paste the key.
Leave Base URL at https://api.greenpt.ai/v1, or set it to https://api.us.greenpt.ai/v1 to keep requests in the US region.
Tested with n8n 2.42.6.
The chat model calls GreenPT's Chat Completions endpoint, so tool calling works with the AI Agent node. You can also set any model with an expression, for example to pick a model per item.