Playground

Test your prompts against AI models and compare results. Open the Playground by clicking Playground in the editor's Actions tab.

The Playground uses a grid layout where prompt versions are rows and AI models are columns. You can run individual cells, entire rows or columns, or the full grid at once.

In this section

• Model & Settings — configure models with temperature, max tokens, and reasoning effort.
• Batch Comparison — add multiple models and versions to the grid for side-by-side comparison.
• AI Judge Scoring — automatic scoring of responses on accuracy, helpfulness, relevance, coherence, and safety.
• Run History — browse past test runs and copy responses.