Nanobanana DL
Project created to generate batch images with Nano Banana, based on virtual gift idea lists. The focus was productivity and results: automate sequential image creation using the free Gemini API, with operation via VS Code and CLI.
Create a fast workflow to generate many coherent images from prompts, without additional cost, with sequence control and file organization. The project needed to work for internal use, without a public interface, prioritizing execution speed and result predictability.
Node automation to consume Gemini via Google AI Studio, with batch prompt pipeline. The workflow uses idea lists (virtual gifts) and generates images sequentially, with name and folder control, maintaining traceability. Operations were performed both in VS Code and via CLI to accelerate testing and execution.
Defined the workflow strategy, wrote prompts and automation, integrated with Gemini API, and validated image quality. Conducted tests, adjusted parameters, and documented operations for quick use in other projects.
Delivery of a functional batch generator, reducing image production time and eliminating manual steps. Using the free API maintained zero cost, and the sequential workflow ensured consistency and speed to feed another project.
Batch prompt engineering requires standardization of structure and parameters, plus naming conventions to avoid rework. The combination of VS Code and CLI brought agility to iterate and compare results.