💙 Auto Discount (10%)

Automatically applies a 10% discount to your usage across all available models and tools.

⚙️ Configure Tools - Complete Guide

The "Configure Tools" dialog allows you to select which capabilities Copilot should have access to. This is critical for security, performance, and workflow optimization.

⚠️ Important: Enabling tools Copilot doesn't need for your task can slow down responses. Disable unused tools for optimal performance.

When & Why to Enable/Disable Each Tool:

✅ Built-In (Always Enable)

Function: Core VS Code integration providing fundamental Copilot functionality within the editor environment.

Enable When:

• Working in VS Code (always - this is required)
• Using any Copilot feature

Disable When:

• Never disable this - it's essential

🤝 Agent - Delegate Tasks to Other Agents

Function: Distributes tasks across multiple AI agents for parallel processing and specialized handling.

Enable When:

• Working on large, multi-faceted projects
• Need parallel task execution
• Different components need different expertise
• You want autonomous agents to handle subtasks

Disable When:

• Working on simple, single-focused tasks
• You need sequential, controlled workflow
• Performance is critical (adds latency)
• Working on sensitive security code

✏️ Edit - Edit Files in Your Workspace

Function: Allows Copilot to directly modify files, enabling automated code generation and refactoring.

Enable When:

• You want Copilot to auto-generate or refactor code
• Performing bulk file modifications
• Creating new files automatically
• You trust Copilot's code changes
• Interactive development workflow

Disable When:

• Review is critical before changes
• Working with production code directly
• You only want suggestions, not automatic changes
• Tight version control requirements

▶️ Execute - Execute Code & Applications

Function: Allows running commands, scripts, and applications on your machine.

Enable When:

• You want Copilot to test generated code
• Running build scripts automatically
• Testing deployments
• Real-time code validation needed
• Development and testing phases

Disable When:

• ⚠️ SECURITY RISK: Public/shared machines
• Production environment (manual execution only)
• You only want suggestions without execution
• Testing untrusted code
• Limited system resources

📖 Read - Read Files in Your Workspace

Function: Enables Copilot to access and analyze files for context-aware assistance.

Enable When:

• You want contextual code analysis
• Copilot needs to understand your codebase
• Code review and refactoring tasks
• Pattern identification needed
• Almost always (very useful)

Disable When:

• Discussing proprietary/confidential algorithms
• Privacy constraints
• Only doing general programming advice
• Very large codebase (performance concern)

🔍 Search - Search Files in Your Workspace

Function: Provides workspace search to locate specific files and patterns.

Enable When:

• Finding implementations across codebase
• Locating usage patterns
• Navigation and discovery needed
• Understanding project structure
• Refactoring across multiple files

Disable When:

• Privacy/confidentiality concerns
• Not working on existing codebase
• Generic programming discussion

✓ Todo - Manage Task Lists

Function: Allows Copilot to create and manage todo lists for project planning.

Enable When:

• Managing complex projects
• Breaking down large features into steps
• Tracking progress on multi-step tasks
• Task coordination
• Sprint planning with Copilot

Disable When:

• Simple, single-task work
• Using external project management tools
• Not planning work with Copilot

🎯 VS Code - Native VS Code Features

Function: Direct access to VS Code commands and APIs for automation.

Enable When:

• You want Copilot to navigate and interact with editor
• Opening files, views, and panels automatically
• Running VS Code commands
• Most development scenarios

Disable When:

• Manual editor control preferred
• Very limited editor interaction needed

🌐 Web - Fetch Information from Web

Function: Retrieves information from external web sources.

Enable When:

• Looking up API documentation
• Research and references needed
• Real-time information required
• Learning about new libraries/frameworks
• Fetching external data

Disable When:

• Air-gapped/offline development
• Security/firewall restrictions
• Privacy concerns about browsing
• Offline documentation available locally

☁️ Azure Resources

Function: Integration with Microsoft Azure cloud services.

Enable When:

• Using Azure for cloud deployment
• Managing Azure resources from Copilot
• Cloud infrastructure automation
• Azure-based projects

Disable When:

• Not using Azure
• On-premise or other cloud providers
• Manual cloud management preferred

📓 Jupyter

Function: Support for Jupyter notebooks with interactive execution.

Enable When:

• Data science and ML projects
• Exploratory analysis work
• Scientific computing
• Interactive visualization
• Research and experiments

Disable When:

• Not using notebooks
• Pure backend/web development
• Not doing data analysis

🎨 Mermaid Chat Features

Function: Create and manage diagrams with Mermaid.

Enable When:

• Creating architecture diagrams
• Flowchart generation
• Documentation with visuals
• System design discussion

Disable When:

• Using other diagram tools
• No need for visual documentation

🐍 Python

Function: Native Python script execution and development support.

Enable When:

• Python development projects
• Script execution and testing
• Automation tasks
• Data processing

Disable When:

• Not using Python
• Using other programming languages only

💡 Pro Tip - Recommended Default Configuration for Developers:

• ✅ Built-In - Always ON
• ✅ Read - ON (for context)
• ✅ Search - ON (for navigation)
• ✅ VS Code - ON (for automation)
• ⚠️ Edit - ON (use with review)
• ⚠️ Execute - CONDITIONAL (only in dev environment)
• ❓ Others - Based on your specific needs

🤖 Claude Models - Complete Analysis

Claude models are powerful AI assistants created by Anthropic, optimized for different use cases with varying capabilities and costs. Choose based on your speed/quality needs and budget.

📊 Cost Index Explanation:

• 0x = Free/minimal cost (best for high-volume tasks)
• 0.33x = ~33% of standard cost (good balance)
• 1x = Standard cost (professional default)
• 3x = 3x standard cost (best quality, premium reasoning)

Models Overview

GPT-4.1 - 0x (FREE)

Speed: Very Fast | Intelligence: Basic | Context: Limited

When to Use:

• Simple text generation or scripts
• Basic coding questions
• Budget is absolute priority
• Testing/prototyping with high frequency
• Learning or experimentation

GPT-4o - 0x (FREE)

Speed: Fast | Intelligence: Good | Cost/Value: Excellent

When to Use:

• General-purpose coding assistance
• Code review and analysis
• Documentation generation
• Standard development tasks
• When budget is tight but quality needed

GPT-4o mini - 0x (FREE)

Speed: Fastest | Intelligence: Adequate | Latency: Minimal

When to Use:

• Real-time completions needed
• High frequency of requests
• Quick iterations and prototyping
• Mobile or edge devices
• Simple to moderate complexity tasks

GPT-5 mini - 0x (FREE)

Speed: Very Fast | Intelligence: Good+ | Value: Strong

When to Use:

• Better reasoning than 4o needed
• Resource-constrained environments
• Latest technology at no cost
• Balancing speed and intelligence

Raptor mini (Preview) - 0x (EXPERIMENTAL)

Status: PREVIEW - Experimental model

When to Use:

• Beta testing new capabilities
• Experimental workflows
• Performance benchmarking
• Research and innovation projects
• ⚠️ Not for production use

💚 Claude Haiku 4.5 - 0.33x (RECOMMENDED)

Status: CURRENT - The model currently being used

Speed: Very Fast ⚡ | Intelligence: Strong | Context: 200K tokens | Cost Factor: 1/3 of standard

When to Use:

• Default choice for most developers
• Real-time code completion
• Quick debugging and refactoring
• Daily development work
• Budget-conscious organizations
• Teams wanting speed without sacrificing quality
• CI/CD pipelines and automation

💰 Economy Analysis: 67% cheaper than standard models while maintaining excellent quality. Best value proposition. ROI: 4-5x better than premium models for typical development tasks.

Claude Opus 4.5 - 3x (PREMIUM)

Speed: Moderate | Intelligence: Highest 🏆 | Context: 200K tokens | Accuracy: 95%+

When to Use:

• Complex algorithm design
• Advanced system architecture
• Business-critical systems
• Deep technical analysis required
• One-off expensive tasks

💰 Economy Analysis: 3x cost = higher accuracy (5-10% better). Use only when ROI justifies premium cost.

Claude Opus 4.6 - 3x (PREMIUM)

Speed: Moderate | Intelligence: State-of-art 🚀 | Context: 200K tokens | Reliability: Highest

When to Use:

• Mission-critical applications
• Advanced research projects
• Complex multi-step reasoning
• High-stakes business decisions
• Where accuracy is paramount

Claude Sonnet 4 - 1x (STANDARD)

Speed: Fast | Intelligence: High (3x Haiku) | Context: 200K tokens | Value: Strong

When to Use:

• Production code generation
• Comprehensive code reviews
• Technical documentation
• Professional team projects
• Enterprise applications

Claude Sonnet 4.5 - 1x (STANDARD)

Speed: Fast+ | Reasoning: Improved | Performance: 10% better

When to Use:

• Daily coding tasks
• Team collaboration projects
• Standard development work
• Teaching and learning
• Update from Sonnet 4

⭐ Claude Sonnet 4.6 - 1x (STANDARD)

Speed: Fastest Sonnet | Intelligence: Highest Sonnet | Value: OPTIMAL ⭐

When to Use:

• General development work (all types)
• Most professional tasks
• Best value at 1x cost
• Teams wanting reliability + speed balance
• Consider vs Haiku 4.5 for your needs

💰 Economy Analysis: 3x cost of Haiku, but 25% better quality. Choice between Haiku 4.5 and Sonnet 4.6 depends on speed vs quality preference for your team.

Gemini 2.5 Pro - 1x (STANDARD)

When to Use:

• Image understanding needed
• Multimodal tasks (text + images)
• Visual code analysis
• Complex problem solving
• Cross-domain applications

Gemini 3 Flash (Preview) - 0.33x (FAST)

Status: PREVIEW - Speed-optimized for high-frequency tasks

When to Use:

• Real-time inference needed
• High-volume query processing
• Beta feature testing
• Speed is priority

Gemini 3 Pro (Preview) - 1x (STANDARD)

Status: PREVIEW - Professional-grade with balanced capabilities

When to Use:

• Complex reasoning tasks
• Advanced analysis needed
• Research applications
• Emerging use cases

Gemini 3.1 Pro (Preview) - 1x (CUTTING EDGE)

Status: PREVIEW - Latest iteration, experimental capabilities

When to Use:

• Next-generation capabilities needed
• Experimentation and innovation
• Research projects
• Future-proofing architecture

GPT-5.1 - 1x (LATEST)

When to Use:

• Advanced development
• Complex problem solving
• Latest technology features
• Frontier applications

GPT-5.1-Codex - 1x (CODE SPECIALIST)

Specialty: Optimized specifically for code generation and analysis

When to Use:

• Code generation quality is critical
• Programming education
• Technical debt resolution
• Code optimization and refactoring
• Cross-language development

🚀 Best for Daily Development

Claude Sonnet 4.6

Best overall value with excellent speed and reasoning. Recommended default choice for most developers.

⚡ Best for Speed

Claude Haiku 4.5

Fastest available model, perfect for quick iterations and real-time completions.

🧠 Best for Complex Tasks

Claude Opus 4.6

Superior reasoning for mission-critical code and advanced problem-solving.

💚 Best for Budget

GPT-4o or GPT-4o mini

Cost-effective options with solid capabilities for standard tasks.

🎨 Best for Multimodal

Gemini 2.5 Pro

Excellent for image understanding and multimodal tasks.

🔬 Best for Experimentation

Preview Models (Raptor mini, Gemini 3.x)

Latest experimental features for innovative projects.

✨ Key Features & Context

📊 Context Windows

Most modern Claude models feature a 200K token context window, enabling:

• Analysis of large codebases
• Deep contextual understanding
• Multi-file refactoring
• Comprehensive documentation

🔧 Tool Ecosystem

The comprehensive tool set enables Copilot to:

• Read and modify your codebase
• Execute code and tests
• Search across files
• Fetch external resources
• Create visual diagrams
• Manage cloud resources

💳 Cost Optimization

The "Auto" discount button provides automatic 10% savings across all models and tools, making premium features more accessible.