Margarita turns cocktail making into a small first-person skill game. Pick your ingredients, choose your own amounts, add ice, shake whenever you want, and serve when you think the Margarita is ready. The game tracks your choices, dilution, temperature, shaking, and mistakes, then scores the final drink from 1–10. It uses interactive 3D bar tools, real-world-inspired ingredients, touch controls, and device-motion shaking on mobile.
What real task does your product handle with GPT-6 Astra?
Maker
GPT-6 Astra helped build the interactive bartender experience end-to-end: creating and refining Blender-authored 3D assets, integrating them into the web scene, and improving the gameplay through repeated visual feedback.
In the product itself, the player freely chooses ingredients, amounts, shaking, dilution, and timing, and the experience evaluates those choices into a final Margarita score instead of forcing a fixed recipe.
The current launch focuses on Margarita, but I plan to expand the same system to more cocktails, each with its own ingredients, interactions, scoring, and physical rituals.
Report
Maker
📌
I built Margarita for the GPT-6 Astra Challenge.
The idea started as a cinematic drink-making experience, but I wanted the user to actually make the drink rather than just watch an animation.
So it became a small first-person bartender game: choose your ingredients, decide the amounts, add ice, shake as much or as little as you want, and serve whenever you think the Margarita is ready. The final drink gets scored from 1–10 based on what you actually made.
I kept the scope intentionally small and focused on one drink, physical interaction, mobile motion controls, and repeated visual/gameplay refinement.
Would love feedback on the interaction and what drink I should build next.