This n8n workflow automates text generation using the IBM Granite Speech 3.3 model hosted on Replicate. It starts with a manual trigger, where the user clicks to execute the process. The workflow first sets the Replicate API key, then sends a POST request to initiate a text prediction. It extracts the prediction ID from the response, waits briefly, and regularly polls the prediction status. Once the prediction succeeds, it processes and outputs the generated text, making it suitable for automating content creation, chatbots, or AI-powered writing tasks. This workflow is ideal for developers or content creators wanting to streamline AI text generation within their automation ecosystem.