Released on July 24, 2026, the latest patch release for n8n, version 2.32.4, brings two important fixes for users leveraging the platform’s AI capabilities. First, the AI Assistant’s verification step now properly falls back to the model specified in the workflow run when the default configuration is unavailable. Previously, if the designated model wasn’t accessible, the verification call could fail, potentially interrupting assistant tasks. Now, the system intelligently resorts to the model resolved for the current run, ensuring smoother and more resilient AI interactions.
Second, a fix prevents Instance AI follow-up runs from entering an infinite loop when failures occur before the agent even starts. In earlier versions, a failing preparation step could trigger repeated, unsuccessful runs, wasting resources and cluttering execution logs. With this update, the platform detects the early failure and halts the loop, saving self-hosted users from unnecessary troubleshooting and resource drain.
