AI Tools

Choosing the Right AI Model

Claude vs GPT vs Gemini vs open-source — a practical guide to picking the right tool

There are dozens of AI tools out there, and it's hard to know where to start. This guide cuts through the noise: here's what each major AI model is good at, what it costs, and when to use which one.

First, what's an AI model?

When you use ChatGPT, Claude, or Gemini, you're talking to an AI model — a program that's been trained on massive amounts of text to understand and generate language. Different companies build different models, and they each have strengths. Think of it like cars: they all drive, but a pickup truck and a sports car are good at different things.

The Big Three

These are the AI models most people interact with. All three are excellent — the differences are in their specialties.

Claude (by Anthropic) — Especially strong at writing, analysis, coding, and working with long documents. Tends to follow complex, multi-step instructions carefully. Available at claude.ai.
ChatGPT (by OpenAI) — The most well-known AI. Great at creative tasks, and can handle text, images, and voice in the same conversation. Huge ecosystem of plugins. Available at chatgpt.com.
Gemini (by Google) — Deep integration with Google products (Docs, Gmail, Search). Has an enormous context window (the amount of text it can "remember" in a single conversation). Available at gemini.google.com.

All three have free tiers. You can try each one without paying anything.

What about open-source models?

The models above all run on company servers — you send your text to their computers, and they send a response back. Open-source models are different: they're free to download and run on your own computer. Nobody else sees your data.

Llama (by Meta) — The most popular open-source model family
Mistral (by Mistral AI) — Fast and capable, great for coding
Qwen (by Alibaba) — Strong multilingual support
Gemma (by Google) — Lightweight, runs well on modest hardware

Open-source models are less capable than the Big Three, but they're free, private, and work offline. See Guide #2 to learn how to run one.

So which one should I use?

Here's a simple cheat sheet. When in doubt, start with whichever one you already have an account for — you can always switch later.

Writing essays, emails, or reports → Claude
Analyzing images or working with audio → ChatGPT
Research or anything connected to Google Workspace → Gemini
Privacy-sensitive tasks (medical, legal, personal) → Local open-source
Coding → Claude Code (see Guide #3) or Cursor

Don't overthink it

The AI landscape changes fast — what's "best" today might not be next month. The most important thing isn't which model you pick, it's how well you communicate with it. A great prompt on any model beats a lazy prompt on the "best" one.

Most people settle on one model for daily use and occasionally try others for specific tasks. That's a perfectly good strategy.

Try this now

Pick one task you do regularly — summarizing notes, drafting an email, brainstorming ideas. Try it on Claude, ChatGPT, and Gemini. Which response do you like best? That preference is worth paying attention to.

All Guides Guide 2 →