Skip to content

modelDecisionTree.md

Latest commit

 

History

History
149 lines (111 loc) · 3.96 KB

modelDecisionTree.md

File metadata and controls

149 lines (111 loc) · 3.96 KB

🔮 GitHub Copilot Model Selection Cheat Sheet

A simple, reliable guide for choosing the right model inside GitHub Copilot (Chat, Inline, or Agents).

Disclaimer

Model availability depends on your Copilot plan and org settings.

flowchart TB
    Start["Start"] --> NeedFast["Need speed & low cost?"]

    NeedFast -- Yes --> FastModels["Fast models"]
    NeedFast -- No --> GeneralCoding["Everyday coding task?"]

    GeneralCoding -- Yes --> BalancedModels["Balanced models"]
    GeneralCoding -- No --> DeepReasoning["Hard problem or deep reasoning?"]

    DeepReasoning -- Yes --> DeepModels["Deep-reasoning models"]
    DeepReasoning -- No --> CodeFocus["Large or strict code generation?"]

    CodeFocus -- Yes --> CodeModels["Code-focused models"]
    CodeFocus -- No --> AgentWorkflows["Using agents or tools?"]

    AgentWorkflows -- Yes --> AgentModels["Agent-strong models"]
    AgentWorkflows -- No --> BalancedModels

    FastModels --> Done["Done"]
    BalancedModels --> Done
    DeepModels --> Done
    CodeModels --> Done
    AgentModels --> Done

    FastModels:::Peach
    DeepModels:::Sky
    CodeModels:::Rose
    AgentModels:::Aqua
    BalancedModels:::Pine

    classDef Peach stroke-width:1px, stroke:#FBB35A, fill:#FFEFDB, color:#8F632D
    classDef Rose stroke-width:1px, stroke:#FF5978, fill:#FFDFE5, color:#8E2236
    classDef Aqua stroke-width:1px, stroke:#46EDC8, fill:#DEFFF8, color:#378E7A
    classDef Sky stroke-width:1px, stroke:#374D7C, fill:#E2EBFF, color:#374D7C
    classDef Pine stroke-width:1px, stroke:#254336, fill:#27654A, color:#FFFFFF
Loading

🧭 Decision Flow

1️⃣ Do you need maximum speed and lowest cost?

➡️ Use FAST models

Models

  • GPT-5 mini
  • Claude Haiku 4.5
  • Gemini Flash

Use when

  • You want quick answers
  • You're summarizing, drafting, or doing simple refactors
  • Low cost matters for large volumes of prompts

2️⃣ Do you need general, everyday coding assistance?

➡️ Use BALANCED models

Models

  • GPT-4.1
  • Claude Sonnet 4 / 4.5
  • Gemini Pro

Use when

  • Writing code
  • Debugging normal issues
  • Explaining code
  • Typical daily work

3️⃣ Are you dealing with a hard problem requiring deep reasoning?

➡️ Use DEEP-REASONING models

Models

  • GPT-5 / 5.1 / 5.2
  • Claude Opus 4.5
  • Gemini 2.5 Pro

Use when

  • Multi-file reasoning
  • Architecture-level analysis
  • Tough bugs
  • Complex logic or multi-step problems

4️⃣ Is the task code-first or code-heavy?

(refactoring, translating, generating large code blocks)

➡️ Use CODE-FOCUSED models

Models

  • GPT-5 Codex
  • GPT-5.1 Codex
  • Codex Mini (preview)

Use when

  • Strict, deterministic code generation
  • Large refactors
  • Code migration across languages

5️⃣ Do you need multi-file edits, tools, or agent workflows?

➡️ Use AGENT-STRONG models

Models

  • Claude Sonnet 4.5
  • GPT-4.1
  • GPT-5 series (if enabled)

Use when

  • Using GitHub Copilot Agents
  • Multi-file refactoring
  • Tool-based modifications
  • Complex workflows

🗂️ Model Family Overview

Category Best For Models
FAST Speed + low cost GPT-5 mini, Claude Haiku 4.5, Gemini Flash
BALANCED Daily coding GPT-4.1, Claude Sonnet 4 / 4.5, Gemini Pro
DEEP-REASONING Complex problem solving GPT-5.x, Claude Opus 4.5, Gemini 2.5 Pro
CODE-FOCUSED Deterministic, strict code GPT-5 Codex, GPT-5.1 Codex, Codex Mini
AGENT-STRONG Multi-file, tool use, agents Claude Sonnet 4.5, GPT-4.1, GPT-5

🎯 Quick Recommendations

  • Most users, most of the time: Claude Sonnet 4.5 or GPT-4.1
  • Hard problems: Claude Opus 4.5 or GPT-5.x
  • Fast & cheap: GPT-5 mini
  • Strict coding: GPT-5 Codex
  • Agent workflows: Claude Sonnet 4.5 or GPT-5