Available AI Models
Different AI models power different features. Learn about available models and when to use each.
Model Overview
What Are Models?
AI models are the engines behind AI features. Different models have different:
- Capabilities
- Speed
- Cost
- Specializations
Available Models
| Model | Best For | Speed | Cost | |-------|----------|-------|------| | GPT-4o | Complex tasks | Medium | Higher | | GPT-4o-mini | Fast simple tasks | Fast | Lower | | Claude 3.5 Sonnet | Long content, analysis | Medium | Medium | | Claude 3 Haiku | Quick responses | Fast | Lower |
Model Capabilities
GPT-4o
Strengths:
- Complex reasoning
- Code generation
- Multi-step logic
- Detailed analysis
Best for:
- Workflow generation
- API analysis
- Complex forms
- Debugging
GPT-4o-mini
Strengths:
- Fast responses
- Cost effective
- Good general use
- Simple tasks
Best for:
- Quick generations
- Simple forms
- Text processing
- Basic automation
Claude 3.5 Sonnet
Strengths:
- Long context
- Nuanced understanding
- Detailed explanations
- Code analysis
Best for:
- Large document analysis
- Complex instructions
- Detailed generation
- Technical content
Claude 3 Haiku
Strengths:
- Very fast
- Cost effective
- Good for simple tasks
- Low latency
Best for:
- Real-time features
- Simple completions
- High-volume tasks
- Quick validations
Choosing a Model
Consider Your Task
| Task Type | Recommended Model | |-----------|-------------------| | Complex workflow | GPT-4o or Sonnet | | Simple form | GPT-4o-mini or Haiku | | API documentation | Sonnet | | Quick generation | Haiku or Mini | | Debugging | GPT-4o | | Large context | Sonnet |
Consider Speed
If speed matters most:
- Claude 3 Haiku (fastest)
- GPT-4o-mini
- Claude 3.5 Sonnet
- GPT-4o
Consider Cost
If budget matters most:
- Claude 3 Haiku (cheapest)
- GPT-4o-mini
- Claude 3.5 Sonnet
- GPT-4o
Default Models
By Feature
| Feature | Default Model | |---------|---------------| | Workflow generation | GPT-4o | | Form generation | GPT-4o-mini | | API import | Claude 3.5 Sonnet | | Quick suggestions | Claude 3 Haiku |
Changing Defaults
To change your default:
- Go to Settings > AI Settings
- Select feature
- Choose preferred model
- Save
Model Settings
Global Settings
Set workspace defaults:
- Default model for each feature
- Fallback model
- Cost limits
Per-Task Override
Override for specific tasks:
- Start AI feature
- Click model selector
- Choose different model
- Generate
Model Comparison
Quality vs Speed
Quality ──────────────────────► Speed
GPT-4o Sonnet Mini Haiku
Complexity vs Cost
| Task Complexity | Recommended | |-----------------|-------------| | Very complex | GPT-4o (worth the cost) | | Complex | Sonnet (balanced) | | Moderate | Mini (cost effective) | | Simple | Haiku (fastest/cheapest) |
Beta Models
Trying New Models
New models appear in beta:
- May have limited availability
- Could have different pricing
- Opt-in to try
Beta Access
- Go to Settings > AI Settings
- Enable Show beta models
- New models appear in selector
Model Updates
Automatic Updates
Models improve over time:
- Better responses
- Faster performance
- Same API interface
Version Pinning
For consistency:
- Some features allow version pinning
- Keeps same model version
- Opt-in setting
Privacy & Security
Data Handling
All models:
- Don't store your data
- Process in real-time
- Follow privacy policy
- Enterprise-grade security
Model-Specific Policies
- OpenAI models: OpenAI's data policy
- Claude models: Anthropic's data policy
- We don't train on your data
Rate Limits
Per-Model Limits
| Model | Requests/min | Tokens/min | |-------|--------------|------------| | GPT-4o | 30 | 40,000 | | GPT-4o-mini | 60 | 60,000 | | Sonnet | 30 | 40,000 | | Haiku | 100 | 100,000 |
Limits vary by plan tier.
Next Steps
- AI Overview - All features
- Credits & Usage - Cost tracking
- Generate Workflows - Use AI