Prompt Outputs: Multi-Model Comparison

Every prompt can keep more than just its text - it can hold the actual results you got back. Attach the answers from ChatGPT, Claude, Gemini and more to a single prompt, each labelled with the model that produced it, and your library turns into a record of which model did the best job. These saved outputs also feed the media walls and the Chrome extension's capture loop.

Multiple Outputs per Prompt

Outputs are stored separately from the prompt, with a one-to-many link, so a single prompt can carry as many results as you like. Each output records the model that generated it, the output type, the content, optional notes and an optional rating from 1 to 10, so you can score and annotate every attempt. To add one, open a prompt and save the result against the model you used - or let the Chrome extension capture it for you. Outputs show up oldest-first, only the owner can manage them, and adding one refreshes the prompt's last-used timestamp.

• Each output stores: the model name, output type, content, optional notes and an optional rating from 1 to 10
• Output types: text, image and video (audio exists in the data model but has no capture UI yet)
• Output content can be very long - up to 500,000 characters
• Many outputs per prompt, shown oldest-first, with owner-only add, edit and delete
• Adding an output updates the prompt's last-used time and logs an 'output_added' entry
• Saved image and video outputs feed the Photo, Video and Media walls