/ Documentation

AI Model Selection & Pricing

Veila gives you access to cutting-edge AI models from multiple providers. Learn how to choose the right model for your tasks, understand pricing, and optimize your usage.

Available AI Models

OpenAI Models

GPT-4o

  • Best for: Complex reasoning, creative writing, detailed analysis
  • Strengths: Most capable model, excellent at following complex instructions
  • Use cases: Research, writing, coding, complex problem-solving
  • Cost: Higher tier pricing

GPT-4o-mini

  • Best for: Most everyday tasks with great capability-to-cost ratio
  • Strengths: Fast, reliable, good reasoning at lower cost
  • Use cases: General questions, coding help, writing assistance, daily tasks
  • Cost: Mid-tier pricing
  • ⭐ Recommended: Great starting point for new users

GPT-3.5-turbo

  • Best for: Simple questions, quick tasks, high-volume usage
  • Strengths: Very fast responses, most economical
  • Use cases: Basic Q&A, simple coding, quick explanations
  • Cost: Most economical option

Anthropic Models

Claude-3.5-Sonnet

  • Best for: Analysis, reasoning, safety-conscious responses
  • Strengths: Excellent at nuanced thinking, ethical considerations
  • Use cases: Research analysis, content moderation, complex reasoning
  • Cost: Premium pricing

Claude-3-Haiku

  • Best for: Fast, cost-effective responses
  • Strengths: Quick processing, good for simple tasks
  • Use cases: Basic questions, simple analysis, quick tasks
  • Cost: Lower-tier pricing

Other Providers

  • Mistral Models: Coming soon with competitive pricing
  • Meta Models: Llama variants available for specific use cases
  • More providers: Regular additions of new models and providers

Model Selection Strategy

Choose by Task Type

Creative Writing & Content

Recommended: GPT-4o > Claude-3.5-Sonnet > GPT-4o-mini
- Long-form content creation
- Story writing and creative projects  
- Marketing copy and blogs
- Poetry and creative expression

Code & Technical Help

Recommended: GPT-4o-mini > GPT-4o > Claude-3.5-Sonnet
- Debugging and code review
- Writing functions and scripts
- Technical explanations
- API documentation help

Analysis & Research

Recommended: Claude-3.5-Sonnet > GPT-4o > GPT-4o-mini
- Data analysis and interpretation
- Research synthesis
- Complex reasoning tasks
- Academic and professional analysis

Quick Questions & Daily Tasks

Recommended: GPT-4o-mini > GPT-3.5-turbo > Claude-3-Haiku
- Simple Q&A
- Quick explanations
- Daily task planning
- Basic calculations and conversions

Cost-Optimization Tips

Start Small, Scale Up

  1. Begin with GPT-4o-mini for most tasks
  2. Switch to premium models only when you need advanced capabilities
  3. Use GPT-3.5-turbo for simple, repetitive questions
  4. Reserve GPT-4o for your most important or complex tasks

Model Switching Strategy

  • Start conversations with cost-effective models
  • Switch mid-conversation when you need more capability
  • Ask simple questions first, then dive deeper with premium models
  • Use context building - let cheaper models set up, then switch for analysis

Understanding Pricing

Pricing Display Options

Veila offers two ways to view pricing:

Intuitive Pricing (Recommended)

  • Shows cost per average A4 page of text (~500 tokens)
  • Easier to understand for non-technical users
  • Real-world reference you can relate to
  • Example: "0.002 / A4 page" means 2/10th of a cent per page

Technical Pricing

  • Shows cost per 1 million tokens
  • Industry standard format
  • Includes cached input pricing for advanced users
  • Example: "2.50 / 1M tokens"

You can toggle between these views in your Settings.

Cost Components

Input Costs (Your Messages)

  • What you pay for sending messages to the AI
  • Length-based: Longer messages cost more
  • Context included: Previous conversation counts toward input

Output Costs (AI Responses)

  • What you pay for AI-generated responses
  • Usually higher than input costs
  • Length varies: AI response length affects total cost

Cached Input (Technical View Only)

  • Discounted rate for repeated content in technical pricing view
  • Automatic optimization by AI providers
  • Lower cost for context that's been seen before

Live Pricing Information

The chat interface shows real-time pricing for each model:

In the Model Selector

  • Hover over any model to see its pricing
  • Compare costs before making your selection
  • Updated pricing reflects current rates + markup

In the Chat Input Area

  • Current model pricing displayed below the input
  • Input and output rates clearly shown
  • Toggle intuitive/technical pricing views
  • "Info" tooltip explains our markup policy

Understanding Our Markup

Veila adds a transparent markup to cover operational costs:

Pricing Policy: We charge the official prices from AI providers (OpenAI, Anthropic, etc.) plus a markup percentage to cover our development and operational costs.

  • Transparent calculation: Base price + markup = your price
  • No hidden fees: Everything is clearly displayed
  • Fair pricing: Markup covers privacy infrastructure and development
  • Updated regularly: Pricing reflects current provider rates

Model Selection in Practice

Switching Models Mid-Conversation

One of Veila's unique features is seamless model switching:

How It Works

  1. Continue any conversation with a different model
  2. Full context preserved - new model sees the entire chat history
  3. Compare responses from different AIs on the same topic
  4. Cost optimization - use expensive models only when needed

Practical Example

You: "Explain quantum computing basics"
[Using GPT-4o-mini]

AI: [Gives basic explanation]

You: [Switch to GPT-4o] "Now explain the technical details of quantum entanglement"
[Using GPT-4o]  

AI: [Gives detailed technical explanation]

Setting Your Default Model

In Chat Interface

  • Model dropdown remembers your last selection
  • Quick switching between your most-used models
  • Per-chat selection - each conversation can use different models

In Settings

  • Set default model for new chats
  • Choose from enabled models only
  • Override anytime in individual conversations

Model Management in Settings

Navigate to Settings > Models to:

Enable/Disable Models

  • Select which models appear in your chat dropdown
  • Hide models you don't use to reduce clutter
  • Access all provider options in one place

Set Default Model

  • Choose your go-to model for new conversations
  • Pick from your enabled models only
  • Change anytime as your needs evolve

Pricing Preferences

  • Toggle intuitive vs. technical pricing
  • Set display preferences that work for you
  • View markup information and transparency details

Model Comparison Tips

Testing Different Models

Same Question, Different Models

Try asking the same question to different models to see:

  • Response style differences
  • Depth of analysis variations
  • Creative approach differences
  • Technical accuracy comparisons

Context Continuation

  • Start with a cheaper model to build context
  • Switch to premium models for deeper analysis
  • Compare how each model interprets the same conversation

Performance Characteristics

Response Speed

  • GPT-3.5-turbo: Fastest responses
  • GPT-4o-mini: Quick responses with better quality
  • GPT-4o & Claude-3.5-Sonnet: Thoughtful, sometimes slower responses

Response Length

  • Claude models: Often more detailed by default
  • GPT models: Concise unless asked for detail
  • Varies by question: Complex topics get longer responses

Accuracy & Reliability

  • Premium models: Higher accuracy on complex topics
  • All models: Reliable for basic information
  • Specialized tasks: Some models excel in specific domains

Troubleshooting Model Issues

Model Not Responding

  1. Check credit balance - insufficient credits prevent responses
  2. Try a different model - temporary provider issues
  3. Refresh the page - clear any interface issues
  4. Check error messages - specific guidance for issues

Unexpected Responses

  1. Try rephrasing your question for clarity
  2. Switch models - different AIs have different strengths
  3. Provide more context - help the AI understand your needs
  4. Break complex questions into smaller parts

Cost Management

  1. Monitor usage in the pricing display
  2. Start with cheaper models for exploration
  3. Use model switching strategically
  4. Check free tier status for daily allowances

Next Steps

Now that you understand model selection:

  1. Learn chat organization - Manage your conversations effectively
  2. Understand credits and billing - Optimize your usage costs
  3. Explore settings - Customize model preferences
  4. Get help and support - Find answers to specific questions

← Chat Basics | Chat Organization →

← Back to Home Need Help? Contact Support →