Hub
Pricing About
Collection

KNIME for Generative AI

AILLMChatModelOpenAI
+1

Explore this collection of ready to use workflows to get started with using Large Language Models (LLMs). Browse workflows for connecting to and interacting with LLMs, building chat apps, performing retrieval augmented generation, and more.

KNIME AI Extension Guide
This webinar guides you through creating your first workflow with the AI extension.
Dec 6, 2023 11:23 AM
Like

Query AI via KNIME

In this section we introduce three Large Language Model (LLM) approaches accessible in KNIME via the KNIME AI Extension (Labs): OpenAI / Azure OpenAI, and Hugging Face. They can be used for text generation, translation, classification, and summarization tasks. Each of these three approaches contains three different nodes: one for authentication with the online model, one for connecting to the online model and one for sending the prompts to the model
OpenAI
Use the latest OpenAI model to classify your documents. OpenAI is most known for its application ChatGPT and offers the most state-of-the-art general models. One will yield the best general results with OpenAI models since they have been trained on a vast amount of data and the most parameters for decision-making.
Node / Source
OpenAI Authenticator

This node provides the authentication for all OpenAI models. Learn more

Node / Source
OpenAI LLM Connector

This node establishes a connection with an OpenAI Large Language Model (LLM). Learn more

Node / Predictor
LLM Prompter

For each row in the input table, the LLM Prompter sends one prompt to the LLM and receives a corresponding response. Learn more

Go to item
Workflow
Simple OpenAI Prompt
OpenAIAIGenerative AI
+1
This workflow shows how to connect to an OpenAI LLM and send prompts to it.
AI Extension Example Workflows1) Large Language Models1.1 - Simple OpenAI Prompt
8
knime
Azure OpenAI
Use the latest OpenAI models that are hosted on the Azure Cloud. These are the same models as the OpenAI ones but hosted in the secure Azure infrastructure.
Node / Source
Azure OpenAI Authenticator

This node provides the authentication for all Azure OpenAI models. Learn more

Node / Source
Azure OpenAI LLM Connector

This node establishes a connection with an Azure OpenAI Large Language Model (LLM). Learn more

Node / Predictor
LLM Prompter

For each row in the input table, the LLM Prompter sends one prompt to the LLM and receives a corresponding response. Learn more

Go to item
Workflow
Simple Azure OpenAI Prompt
OpenAIAILLM
+2
This workflow shows how to connect to an Azure OpenAI LLM and send prompts to it.
AI Extension Example Workflows1) Large Language Models1.6 - Simple Azure OpenAI Prompt
0
knime
Hugging Face
HuggingFace is an open-source community for building applications using machine learning models. Use their free or paid text-generating models right from the Hugging Face hub. These models can produce better results than the ones from OpenAI depending on the use case since HuggingFace offers the possibility to up and download specifically fine-tuned models.
Node / Source
HF Hub Authenticator

This node provides the authentication for all Hugging Face Hub models. Learn more

Node / Source
HF Hub LLM Connector

This node establishes a connection to a specific LLM hosted on the Hugging Face Hub. Learn more

Node / Predictor
LLM Prompter

For each row in the input table, the LLM Prompter sends one prompt to the LLM and receives a corresponding response. Learn more

Go to item
Workflow
Simple HuggingFace Prompt
AIGenerative AILLM
+1
This workflow shows how to connect to an LLM on Hugging Face Hub and send prompts to it.
AI Extension Example Workflows1) Large Language Models1.2 - HuggingFace Hub
4
knime

Build a chatbot

KNIME makes it easy for you to use chat models. Chat models are a subset of LLMs trained for specific conversation inputs. For this reason, it is possible to connect each chat model to a LLM port but not vice versa.
Build with Open AI
Use the ChatGPT Application’s underlying model to build your chat with KNIME.
Node / Source
OpenAI Authenticator

This node provides the authentication for all OpenAI models. Learn more

Node / Source
OpenAI Chat Model Connector

This node establishes a connection with an OpenAI Chat Model. Learn more

Node / Predictor
Chat Model Prompter

The Chat Model Prompter takes a statement (prompt) and the conversation history with human and AI messages. Learn more

Go to item
Workflow
Simple ChatGPT Prompt
ChatGPTChat ModelLLM
+1
This workflow shows how to connect to the model behind ChatGPT and send prompts to it.
AI Extension Example Workflows2) Chat Model2.1 - Simple ChatGPT prompt
2
knime
Go to item
Workflow
ChatGPT Data App
OpenAIAIChatGPT
+1
This workflow shows how to build a simple chatbot data app that is powered by ChatGPT.
AI Extension Example Workflows2) Chat Model2.2 - ChatGPT Data App
4
knime
Build with Azure OpenAI
Use the latest OpenAI models that are hosted on the Azure Cloud. These are the same models as the OpenAI ones but hosted in the secure Azure infrastructure.
Node / Source
Azure OpenAI Authenticator

This node provides the authentication for all Azure OpenAI models. Learn more

Node / Source
Azure OpenAI Chat Model Connector

This node establishes a connection with an Azure OpenAI Chat Model. Learn more

Node / Predictor
Chat Model Prompter

The Chat Model Prompter takes a statement (prompt) and the conversation history with human and AI messages. Learn more

Go to item
Workflow
Azure OpenAI ChatApp
OpenAIAIChatGPT
+3
This workflow shows how to build a simple chatbot data app that is powered by GPT-3.5 Turbo (free version of ChatGPT).
AI Extension Example Workflows2) Chat Model2.4 - Azure OpenAI ChatApp
1
knime
Add related information to your chat
Not only can you create your chatbot with KNIME, but you can also provide it with your data. Cut the cost of training or fine-tuning your model by providing it directly with the necessary content. Preprocess your documents in KNIME, store them in a local Vector Store and retrieve semantically similar documents to your question to enrich the content given to your chatbot.
Node / Source
FAISS Vector Store Creator

The node generates a FAISS vector store by processing a string column containing documents with the provided embeddings model. Learn more

Node / Source
Vector Store Retriever

A vector store retriever is a component that specializes in retrieving vectors from a vector store based on user queries. Learn more

Node / Source
OpenAI Embeddings Connector

This node establishes a connection with an OpenAI Embeddings Model. Learn more

Node / Source
OpenAI Authenticator

This node provides the authentication for all OpenAI models. Learn more

Go to item
Workflow
Create a Vector Store
OpenAIEmbeddingsVector Store
+1
This workflow shows how to create a vector store from a KNIME table using OpenAI Embeddings and the FAISS Vector Store Creator.
AI Extension Example Workflows3) Vector Stores3.1 - Create a Vector Store
2
knime
Information retrieval with an agent
You can further empower your chatbot by turning it into an agent. Since retrieving related information from a vector store is not always necessary, an agent decides whether it needs additional information sources to answer. Providing the agent with a set of different knowledge bases, it will automatically choose whether and from where information will be retrieved.
Node / Source
OpenAI Functions Agent Creator

This node creates an agent based on OpenAI chat models that support function calling. Learn more

Node / Predictor
Agent Prompter

This node combines an agent with a set of tools, and prompts it with a user provided prompt and the conversation history. Learn more

Node / Source
FAISS Vector Store Reader

This node reads a FAISS vector store created with another tool (e.g. Python) from a local path. Learn more

Node / Manipulator
Vector Store to Tool

Turns a vector store into a tool by providing it with a name and a description. Learn more

Node / Source
OpenAI Embeddings Connector

This node establishes a connection with an OpenAI Embeddings Model. Learn more

Node / Source
OpenAI Authenticator

This node provides the authentication for all OpenAI models. Learn more

Node / Source
OpenAI Chat Model Connector

This node establishes a connection with an OpenAI Chat Model. Learn more

Go to item
Workflow
OpenAI Agent
OpenAIAIAgent
+1
This workflow shows how to provide an OpenAI Functions Agent with a vector store as tool.
AI Extension Example Workflows4) AgentsOpenAI Agent
3
knime
Go to item
Workflow
OpenAI Agent ChatApp
OpenAIAIAgent
+1
This workflow uses an OpenAI Functions Agent and multiple vector stores to power an interactive chat dashboard.
AI Extension Example Workflows4) AgentsOpenAI Agent ChatApp
1
knime

KNIME AI Extension (Labs)

Check out the whole extension below. Note that it is still in labs, so there can be compatibility changes because some features are significantly actively developed currently. Please share your feedback in the KNIME forum.
Go to item
Extension
KNIME AI Extension (Labs)Trusted extension
This KNIME extension connects to LLMs, chat and embeddings models from OpenAI, Hugging Face Hub and GPT4ALL.

KNIME AG, Zurich, Switzerland

10
knime
Provide feedback: KNIME Forum

More Resources

Blog: A friendly intro to Large Language Models
Blog: Four levels of LLM customization

KNIME
Open for Innovation

KNIME AG
Talacker 50
8001 Zurich, Switzerland
  • Software
  • Getting started
  • Documentation
  • Courses + Certification
  • Solutions
  • KNIME Hub
  • KNIME Forum
  • Blog
  • Events
  • Partner
  • Developers
  • KNIME Home
  • Careers
  • Contact us
Download KNIME Analytics Platform Read more about KNIME Business Hub
© 2025 KNIME AG. All rights reserved.
  • Trademarks
  • Imprint
  • Privacy
  • Terms & Conditions
  • Data Processing Agreement
  • Credits