Promptura

Iteratively discover the best prompt for your task

What it does

Promptura takes a task description and a model (fal.ai or EachLabs): it generates diverse candidate prompts, runs them on the model, and shows results side by side. You pick the outputs you like and add notes; the next round produces 10 refined prompts based on your feedback. A structured loop to find the right prompt for text, image, and video generation.

Current supported modalities: text-to-image, image-to-image, text-to-video.

  • Configure API keys: In Settings, add at least one provider key (fal.ai or EachLabs). Required before using the Playground.
  • Enter a task: e.g. “A toy car on a wooden table” — describe your goal for text or image generation.
  • Generate candidates: The system produces multiple prompts tailored to your chosen model; each is run and results appear as cards.
  • Select and refine: Choose the ones you like, add notes, and generate the next round of improved prompts.

Add your API keys in Settings first, then open the Playground.

Go to PlaygroundOpen Settings →