AI Models Overview

AI Models Overview

PressPlay integrates advanced AI models to help you generate compelling app store assets. Understanding the capabilities and optimal use cases for each model will help you create more effective content for your app store presence.

Supported Asset Types

PressPlay's AI models can generate the following asset types:

• App Icons (IC) - AI-generated icon designs based on your brand and competitive analysis

• Feature Graphics (FG) - Hero images for your store listing header

• Short Descriptions (SD) - Promotional taglines up to 80 characters

Not supported by AI generation:

• Screenshots (SS) - Manual upload only, but automated order testing available

• Promo Videos (PV) - Manual upload only

• Long Description (LD) - Manual upload only

• Multi-Asset experiments - Manual setup only

For asset types that don't support AI generation, see Manual Image Upload.

Choosing the right AI image model

Selecting the appropriate AI model depends on your specific goals and the type of assets you're generating:

Use Gemini When:

• You need highly creative, engaging content

• You're exploring new messaging angles and brand positioning

• You want to test dramatically different content variations

• You're targeting specific emotional responses or user motivations

• You need content that adapts to different cultural contexts

• You're generating longer-form content like full app descriptions

Use Kontext AI When:

• ASO performance is your primary concern

• You need to optimize for specific keywords

• You're working within strict character limits

• You want content aligned with platform-specific best practices

• You're optimizing existing content rather than exploring new directions

• You need category-specific expertise and industry knowledge

Model Configuration

When generating assets, you can select your preferred AI model from the generation settings. Each model can be fine-tuned with:

• Research Methods: The approach the AI uses to analyze your app and market (see Research Methods)

• Competitor Analysis: Specific apps to analyze for inspiration (see Competitor Analysis)

• Target Keywords: Keywords to emphasize in the generated content

• Brand Voice: Tone and style preferences for your content

• Content Constraints: Character limits, required elements, and formatting requirements

Model Performance and Quality

Both models are continuously updated to improve generation quality. You can evaluate their performance by:

• Reviewing Generated Assets: Compare outputs from different models for the same brief

• Tracking Experiment Results: Monitor which model's content performs better in A/B tests

• Analyzing Conversion Metrics: Track how assets from each model impact download rates

• Gathering Team Feedback: Collect input from your team on content quality and relevance

Combining Models in Your Workflow

Many successful teams use both models strategically:

1. Exploration Phase: Use Gemini to generate diverse creative concepts and messaging angles

2. Refinement Phase: Use Kontext AI to optimize the best concepts for ASO performance

3. Testing Phase: Run experiments comparing outputs from both models

4. Optimization Phase: Use performance data to inform which model to use for future generations

Best Practices

• Test Both Models: Generate assets with both models and compare results before committing to one approach

• Provide Clear Context: The more context you provide about your app and goals, the better both models will perform

• Iterate and Refine: Use generated content as a starting point and refine based on your expertise

• Monitor Performance: Track which model produces better results for different asset types

• Stay Updated: Check for model updates and new capabilities as they're released

Next Steps

• Configure Research Methods to guide AI generation

• Add Competitor Apps for market context

• Start your first Asset Generation Run

• Learn about Reviewing Generated Assets