💡 Discovered a cool tool today for testing GenAI models!

Beyond benchmark scores, GitHub marketplace models lets you intuitively test models directly with your own prompts. For free. Hit the Compare button to see two models tackle the same prompt side-by-side.

Even works with multi-modal models.

You can adjust and sync the system prompts and parameters.

Thanks to John Savill for sharing this tip in his excellent Azure AI Foundry overview.