Model Selection Guide
Choosing the right model for your needs is essential for optimal performance and cost-efficiency. This guide will help you understand the different models available in Your App and how to select the most appropriate one for your use case.
Available Models
Standard
Balanced performance for everyday tasks. Good for general text generation and comprehension.
Advanced
Enhanced capabilities for complex reasoning and specialized tasks. Higher token limit.
Expert
Our most powerful model for tasks requiring deep expertise and nuanced understanding.
Fast
Optimized for speed. Ideal for real-time applications and quick responses.
Efficient
Designed for cost-efficiency. Best for high-volume processing with moderate complexity.
Custom
Fine-tuned models for specific domains or tasks based on your data.
Selection Factors
When choosing a model, consider these key factors:
Task Complexity
- Simple tasks (classification, basic Q&A): Standard or Fast models
- Medium complexity (content generation, summarization): Advanced model
- Complex tasks (research, analysis, specialized domains): Expert model
Response Time Requirements
- Real-time applications: Fast model
- Interactive but not time-critical: Standard model
- Batch processing or background tasks: Advanced or Expert models
Token Limits
Different models have different context window sizes (the amount of text they can process at once).
- Standard: 4,000 tokens
- Advanced: 8,000 tokens
- Expert: 16,000 tokens
- Fast: 2,000 tokens
- Efficient: 4,000 tokens
Cost Considerations
Models are priced per 1,000 tokens processed (input + output):
| Model | Price per 1K tokens |
|---|---|
| Standard | $0.005 |
| Advanced | $0.015 |
| Expert | $0.030 |
| Fast | $0.004 |
| Efficient | $0.003 |
Domain Specificity
If your task involves specialized knowledge (legal, medical, scientific, etc.), consider:
- Using the Expert model, which has broader knowledge across domains
- Fine-tuning a Custom model on your domain-specific data
- Using specialized models if available for your domain
Recommended Configurations
Here are some common use cases and the recommended model configurations:
Content Creation
- Blog posts: Advanced model with temperature 0.7
- Technical documentation: Expert model with temperature 0.2
- Creative writing: Advanced model with temperature 0.9
Conversational AI
- Customer support: Standard model with temperature 0.5
- Virtual assistants: Advanced model with temperature 0.6
- Educational chatbots: Expert model with temperature 0.4
Data Analysis
- Basic data summarization: Standard model with temperature 0.3
- Complex pattern recognition: Expert model with temperature 0.2
- Report generation: Advanced model with temperature 0.4
Testing Your Model Selection
It's recommended to test different models for your specific use case before committing to a particular configuration. You can use our Model Comparison Tool to evaluate performance across different models and parameters.
Start with the simplest (and most cost-effective) model that might meet your needs, and only upgrade if necessary. This approach will help optimize your costs while ensuring adequate performance.