Crescent Remixer is an AI-powered application designed to transform your photos into amazing, new scenes and characters with a single click. Upload a picture of a person, choose some fun options, and let the AI instantly remix your image!
What It Does
This app is a creative playground for your photos, turning a standard portrait into something hilarious and unexpected.
Image Transformation: Easily upload a photo of a person and transform them into scenes like a pirate, an astronaut, a superhero, and more.
Easy Prompting: Use the Prompt Builder to choose from simple categories like role, action, emotion, and clothes to build your creative prompt.
Smart Suggestions: The app provides dynamic prompt suggestions (like “Astronaut,” “singer,” “dancing,” or “wearing glasses”) that change based on what you’ve already selected, making the process intuitive and fun.
How to Get Started
This project contains everything you need to run the application locally.
Prerequisites: You need Node.js installed on your machine.
Install dependencies:
npm install
Set the API Key:
You need a Gemini API Key to run the app. Create a file named .env.local in the root directory and add your key:
.env.local
GEMINI_API_KEY="YOUR_API_KEY_HERE"
Run the app:
npm run dev
The app will start and be available in your web browser, typically at http://localhost:3000.
Under the Hood
The Crescent Remixer app is a React and TypeScript project that leverages Google’s Gemini models for its functionality:
Image Transformation: It uses the gemini-2.5-flash-image model. The prompt specifically requests the model to transform the image while maintaining the person’s facial features.
Contextual Prompts: A separate model call generates the creative suggestions for the prompt builder in real-time, providing an interactive experience.
Cuz all apps made in AI Studio are supposed to be able to be run locally.
Instructions how to run locally are set as the readme automatically when you export a project to Gitub.
Alright, after doing some debugging on local use, I’ve come to the conclusion that your code is correct, but the API usage of this model is not free. Ai studio has its own system.
Also, I noticed the model performs best by selecting the least amount of prompts possible. Too much and the face gets messed up.
I dunno. Tried switching the model to 2.0 flash exp and it does work locally.
With gemini-2.5-flash-image, I get this error:
Remix failed: {"error":{"code":429,"message":"You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.\n* Quota exceeded for metric: generativelanguage.googleapis.com/generate_content_free_tier_input_token_count, limit: 0\n* Quota exceeded for metric: generativelanguage.googleapis.com/generate_content_free_tier_requests, limit: 0\n* Quota exceeded for metric: generativelanguage.googleapis.com/generate_content_free_tier_requests, limit: 0\nPlease retry in 4.763020883s.","status":"RESOURCE_EXHAUSTED","details":[{"@type":"type.googleapis.com/google.rpc.QuotaFailure","violations":[{"quotaMetric":"generativelanguage.googleapis.com/generate_content_free_tier_input_token_count","quotaId":"GenerateContentInputTokensPerModelPerMinute-FreeTier","quotaDimensions":{"location":"global","model":"gemini-2.5-flash-preview-image"}},{"quotaMetric":"generativelanguage.googleapis.com/generate_content_free_tier_requests","quotaId":"GenerateRequestsPerMinutePerProjectPerModel-FreeTier","quotaDimensions":{"model":"gemini-2.5-flash-preview-image","location":"global"}},{"quotaMetric":"generativelanguage.googleapis.com/generate_content_free_tier_requests","quotaId":"GenerateRequestsPerDayPerProjectPerModel-FreeTier","quotaDimensions":{"model":"gemini-2.5-flash-preview-image","location":"global"}}]},{"@type":"type.googleapis.com/google.rpc.Help","links":[{"description":"Learn more about Gemini API quotas","url":"https://ai.google.dev/gemini-api/docs/rate-limits"}]},{"@type":"type.googleapis.com/google.rpc.RetryInfo","retryDelay":"4s"}]}}
