Lab samples are translated from their original Persian outputs; the models were not rerun.

Prompt Lab: Real Model Results

One prompt, multiple models: See and compare real, unedited outputs of AI models on Persian prompts side by side.

Lab workflow

  • Every prompt in the library has an "execution example" with its variables filled.
  • The exact same text without any changes is given to each model, and the response is saved without editing.
  • The exact date and model ID are recorded next to each output to make it reproducible.
  • Current lab models: Claude Opus 5.5. More models will be added gradually.