Credit System for Image and Model Usage
Today's Progress
We're releasing v1.1.0, introducing a credit system to manage usage of image generation and specific models within certain subscription plans.
Key Achievements
This update adds a more granular way to track and control resource consumption:
-
Image Generation Credits: Specific subscription plans now include a set number of credits that can be used for image generation. Each generation consumes a certain number of credits, depending on factors like:
- Image resolution.
- Selected AI model.
-
Model Usage Credits: For plans that offer access to premium or specialized models, a separate credit system is in place to track usage. This allows for:
- Fair allocation of resources for computationally intensive models.
- Clear tracking of model usage within the subscription.
-
Credit Tracking and Display: Users can now view their remaining credits within their account dashboard. This provides:
- Real-time visibility into credit balance.
- Clear information on credit consumption for each generation/model usage.
Technical Improvements
Several technical improvements were made to support the credit system:
- Developed a robust credit tracking and management system within the application's backend.
- Integrated credit consumption logic into the image generation and model usage pipelines.
- Updated the user interface to display credit information clearly and concisely.
Challenges
Implementing the credit system presented some challenges:
- Designing a fair and balanced credit system that reflects the computational cost of different features and models.
- Ensuring accurate tracking of credit consumption to prevent discrepancies.
- Communicating the credit system clearly to users through the user interface and documentation.
Thoughts
This update provides a more transparent and manageable way for users to utilize our image generation and model resources. The credit system allows for greater flexibility in offering different subscription tiers and ensures fair resource allocation across all users.