KNIME Release Notes

You can now see data output by your agent using the new Agent Chat Widget node (experimental). It provides the same interactive experience as the Agent Chat View node and additionally outputs the data produced by your agent through an output port.

You can connect this data to other nodes to build multi-stage data apps or use it for governance and traceability by recording how agents operate and what data they produce. This gives you more continuity and transparency when building agentic solutions.

The enhanced workflow canvas rendering system introduced in version 5.5 as an opt-in feature is now the default.

Your large workflows load faster and respond better with the new rendering system. Ideal for power users, those building enterprise-scale data pipelines, or anyone working regularly with large workflows.

KNIME Pro is a cloud-based plan that bridges the gap between KNIME Analytics Platform and KNIME Team. It gives individual users an easy way to automate workflows, share insights interactively, and speed up workflow building with K-AI (KNIME AI assistant).

Main features:

• Workflow Scheduling: Run workflows automatically on a schedule — daily, weekly, or at custom intervals.
• Data Apps: Deploy workflows as a shareable data app. End users can explore results interactively without needing KNIME Analytics Platform.
• K-AI Assistant: Accelerate workflow building with expanded access to AI-powered suggestions, Python node configuration, visualization generation, etc.
• Versioning: Track workflow changes with the ability to revert or compare
• Secret management: Centrally store and manage logins to other systems
• REST services: Deploy workflows as API endpoints to power your stack (e.g.real-time recommendations, data enrichment, trigger-based automation).

KNIME Pro is available now. Start your free trial today.

The Agent Chat View node now provides a timeline to let you see the agent’s tool calls and reasoning as they happen, live and expandable within the chat interface.

If you work with agentic workflows, it's helpful to see what the agent is doing behind the scenes. The timeline view separates user-agent conversation from internal tool interactions, making it easier for you to follow the agent’s process. Whether you're developing agent-powered data apps or using them, you now have a transparent view to debug and understand the agent’s decisions.

The OpenAI Authenticator node now accepts a Credentials object to connect to other systems with OpenAI-compatible APIs using OAuth. You can also configure custom request headers directly in the node.

If you work in a large enterprise, you need more control over how authentication and API requests are handled, whether for security, compliance, or integration with internal infrastructure. These updates make it easier for you to connect KNIME to OpenAI-compatible services in a way that fits enterprise requirements.

Use the advanced image generation and editing capabilities of the Gemini 2.5 Flash Image model with the new Gemini Image Generator node

You can use this node to generate new images based on text prompts, fuse several images into one, or make minute edits to existing images without losing their original integrity. You can programmatically generate consistent, high-quality images or make detailed edits as a step in your workflow.

The KNIME Reporting extension now supports view nodes built with Python, such as the Python View node or Geospatial View Static node, allowing you to include them in reports.
Note: Only non-interactive view content is supported.

If you build reports, it's now easier for you to include a wider range of visual outputs in them, especially if you already use Python-based custom views.

You can now see data type icons next to columns when making a selection.

Without visible data types, it can be difficult to choose the right column or know what’s expected. If you build KNIME workflows, this update improves clarity and helps you speed up node configuration.

When your agent calls a tool, the tool’s output view is displayed inside the Agent Chat View node. You can see the visualization, making it easier to follow what the tool is doing and how the agent is using it.

Tool calls now appear in the Agent Chat View node in real time. This helps you stay in sync and understand the agent’s process step by step.

The Tool Message Output node now uses the Message data type, which gives every message a clear role (User, AI, or Tool). You’ll see a more consistent output format and clearer separation of content, so it’s easier to trace what the agent said, what it asked the tool to do, and what result the tool produced.

The new minimap gives you a zoomed-out view of your workflow. You can quickly spot nodes outside your current view and move around more easily, helping you navigate large workflows faster and in a more manageable way.

To use the minimap, make sure WebGL rendering is enabled in Preferences:

Preferences → KNIME → KNIME Modern UI → Workflow editor → WebGL rendering