Image generation on Turbofy with FLUX (BlackForestLabs), Gemini, and GPT Image 2.5 Sunburst (OpenAI)

In this episode of Turbofy Your Work Life, Jens demonstrates how he built Image Studio, an AI-powered image generation and editing app connected to Turbofy through the Model Context Protocol (MCP).

The app was developed with a coding agent and Turbofy MCP, showing how AI coding agents can work with a collaborative cloud workspace to build, improve, automate, and host real applications.

Image Studio supports multiple image-generation providers and models, allowing users to choose different sizes, quality settings, and workflows. It also supports reference images for image editing, making it possible to create a new image from a prompt and then refine it with a natural-language instruction.

In the demonstration, a simple prompt is used to create a busy street scene near the Brandenburg Gate in Berlin. The image is then edited into a sunset version, showing how quickly an AI image workflow can move from generation to refinement.

The video also explains how the app’s automation flow works. Image-generation triggers can start when a new image is created or when an existing image is updated. The flow can connect to providers such as OpenAI, Google, and Black Forest Labs, while securely managing API keys through Turbofy secrets.

Turbofy keeps sensitive API keys out of the coding-agent logs, and the connected MCP workflow can inspect logs, identify errors, correct failed steps, and rerun the process until it works. Generated images are also uploaded automatically to workspace storage so they can be reused in websites, presentations, social media posts, and other applications.

This is a practical example of vibe coding, AI-assisted app development, MCP integration, and building an AI app with a coding agent.