How GeraPrompts works
The marketplace for vetted, runnable AI prompts.
Quick answers
- What can I do with GeraPrompts?
- Find prompts and prompt-chains for tasks you do regularly (drafting, research, analysis, code review), run them on the model of your choice, save tuned variants, and share with your team.
- Why is this different from a list of prompts on a blog?
- Each prompt is runnable on the spot, version-controlled, benchmarked, and licensed. You see real outputs and measured quality before you pay. Authors are paid for value, not blog clicks.
- Which models are supported?
- OpenAI GPT-4 family, Anthropic Claude, Google Gemini, Mistral, Llama, and on-device models. Each prompt declares which models it has been tuned and tested for.
- How is a prompt benchmarked?
- Authors publish expected outputs on a fixed input set. We run that on the declared models and publish accuracy, latency, and cost. Buyers see the report before purchase.
The journey, step by step
- 1
Browse prompts
Search by use-case, model, language, rating, or licence. Read sample outputs before you commit.
- 2
Run on your model
Pick GPT-4, Claude, Gemini, or any compatible model. Run inside GeraPrompts or via the API. Iterate with the author's prompt as your starting point.
- 3
Save or share
Save a tuned variant to your library, share a workflow with your team, or list your own version on the marketplace.
Ready to start?
GeraPrompts is a marketplace where authors publish vetted prompts and full prompt-chains, and buyers run them on their own model of choice. Every prompt is benchmarked, version-controlled, and runnabl