---
title: "AI Model Cheat Sheet: Claude, ChatGPT, and Gemini"
canonical: https://www.ianlurie.com/writing/ai-model-cheat-sheet/
date: 2026-05-13
category: "AI"
description: "A post that tells you which feature on which AI platform does what."
---
[Ian Lurie](https://www.ianlurie.com), May 13, 2026

Agents, Gems, GPTs… I sure wish all these platforms used the same vocabulary. For now, though, I’ve tried to put together a page that tells you which feature on which platform does what. When the underlying feature is the same across platforms, the description is the same too — so you can spot at a glance where the platforms genuinely match up.

## Core concepts

<div class="aiv-table-wrap"><table><colgroup><col class="concept-col"> <col class="platform-col"> <col class="platform-col"> <col class="platform-col"></colgroup><thead><tr><th>Concept</th><th class="platform">Claude</th><th class="platform">ChatGPT</th><th class="platform">Gemini</th></tr></thead><tbody><tr><td><div class="concept">Reusable assistant</div><div class="concept-desc">Something repeatable. Think instructions, context, and rules that you can invoke in multiple sessions to produce consistent results.</div></td><td><span class="equiv-name">Skill</span> <span class="equiv-note">A packaged assistant with custom instructions and uploaded resources (optional), reusable across chats.</span></td><td><span class="equiv-name">Custom GPT</span> <span class="equiv-note">A packaged assistant with custom instructions and uploaded resources (optional), reusable across chats.</span></td><td><span class="equiv-name">Gem</span> <span class="equiv-note">A packaged assistant with custom instructions and uploaded resources (optional), reusable across chats.</span></td></tr><tr><td><div class="concept">Agents</div><div class="concept-desc">Agents perform multi-step tasks. This is what most of us think of when we think “AI.” You give it an instruction, and the agent goes and works its magic, returning a result.</div></td><td><span class="equiv-name">Workflow / Cowork</span> <span class="equiv-note">Cowork (a desktop tool for non-developers) or Claude Code (for developers and the nerds among us). Claude Desktop does similar stuff. Drives tasks across files, connectors, and code.</span></td><td><span class="equiv-name">Agent Mode</span> <span class="equiv-note">Fires on a task-by-task basis from the composer. Plans and executes using a virtual browser, code execution, and connectors.</span></td><td><span class="no-equiv-name">No single named feature</span> <span class="equiv-note">The closest thing to “agents” is a combination of <strong>Deep Research</strong> and <strong>Canvas</strong>. You tie them together.</span></td></tr><tr><td><div class="concept">Persistent workspace</div><div class="concept-desc">A set of instructions, files, and a chat history the model can use to carry out a specific type of task without re-explanation.</div></td><td><span class="equiv-name">Project</span> <span class="equiv-note">A workspace with custom instructions, uploaded files, and its own separate chat history. Context stays within the project.</span></td><td><span class="equiv-name">Project</span> <span class="equiv-note">A workspace with custom instructions, uploaded files, and its own separate chat history. Context stays within the project.</span></td><td><span class="equiv-name">Gem + Drive, or NotebookLM</span> <span class="equiv-note">Gemini doesn’t have “Projects” per se. They have a combination of Gems, Drive access, and NotebookLM. Don’t let the multiple tools scare you. It’s very, very powerful.</span></td></tr><tr><td><div class="concept">Custom instructions</div><div class="concept-desc">The “system prompt” we’ve all heard so much about. They apply to every chat, no matter what.</div></td><td><span class="equiv-name">User preferences / Styles</span> <span class="equiv-note">Global instructions about you and how to respond. Applies to every chat unless overridden inside a Skill or Project.</span></td><td><span class="equiv-name">Custom instructions</span> <span class="equiv-note">Global instructions about you and how to respond. Applies to every chat unless overridden inside a Custom GPT or Project.</span></td><td><span class="equiv-name">Saved info</span> <span class="equiv-note">Global instructions about you and how to respond. Applies to every chat unless overridden inside a Gem.</span></td></tr><tr><td><div class="concept">App and service connectors</div><div class="concept-desc">How you connect to other services like Google Drive. Doesn’t include MCP servers or custom API connections.</div></td><td><span class="equiv-name">Connectors</span> <span class="equiv-note">First-party connections to external apps.</span></td><td><span class="equiv-name">Connectors / Apps</span> <span class="equiv-note">First-party connections to external apps.</span></td><td><span class="equiv-name">Workspace integration / Extensions</span> <span class="equiv-note">It’s Google, so native integration with Gmail, Docs, Sheets, Drive, Calendar, and Meet built in. Otherwise, first-party connectors to external apps.</span></td></tr><tr><td><div class="concept">Command-line agents</div><div class="concept-desc">Direct, non-GUI access to the AI. It’s a great nerd flex, and it’s the most powerful way to access the model.</div></td><td><span class="equiv-name">Claude Code</span> <span class="equiv-note">Command-line tool that drives the model from the terminal. Also has IDE integration with VS Code and other tools.</span></td><td><span class="equiv-name">Codex CLI</span> <span class="equiv-note">Command-line tool that drives the model from the terminal.</span></td><td><span class="equiv-name">Gemini Code Assist / Gemini CLI</span> <span class="equiv-note">Command-line tool that drives the model from the terminal. Also has IDE integration with VS Code and other tools.</span></td></tr><tr><td><div class="concept">Project context file (for command-line agents)</div><div class="concept-desc">A markdown file placed in your project root or in the directory tree. The CLI agent reads it before performing a task. This is the CLI agent’s superpower. Learn to use it!</div></td><td><span class="equiv-name">CLAUDE.md</span> <span class="equiv-note">A markdown file the agent loads before it executes any task within that directory tree. Note that Claude will accept AGENTS.md as well.</span></td><td><span class="equiv-name">AGENTS.md</span> <span class="equiv-note">A markdown file the agent loads before it executes any task within that directory tree.</span></td><td><span class="equiv-name">GEMINI.md</span> <span class="equiv-note">A markdown file the agent loads before it executes any task within that directory tree.</span></td></tr></tbody></table></div>

This is how I [teach the shit out of everything](/writing/teach-everything/). Got questions? Reach out to me! Seriously, I love to help out:
