ChatGPT vs Claude vs Gemini: An Actually Honest Comparison
Marcus RodriguezI'm tired of comparison articles that end with "and the winner is ChatGPT!" or whatever model the author happens to prefer. Real usage is messier than that.
I've been paying for ChatGPT Plus, Claude Pro, and Gemini Advanced for over a year now. I use all three daily, for different things. Here's what I've actually learned—no artificial winner, no sponsored takes.
They've developed real personalities
A year ago, you could swap one model for another and barely notice. That's not the case anymore. Each one has developed distinct strengths, weaknesses, and honestly... personalities.
ChatGPT (GPT-5) feels like the smart generalist friend who's good at almost everything but occasionally gets facts wrong and won't admit it.
Claude (Opus 4.1) feels like the thoughtful writer who genuinely considers your question but sometimes overthinks simple requests.
Gemini (Gemini 3 Pro) feels like the tech-savvy researcher who always has the latest info but can be weirdly literal sometimes.
Coding: Claude wins, but Gemini is close
I've done a lot of coding with all three this year. Claude consistently produces the cleanest, most thoughtful code. It asks good clarifying questions. It considers edge cases I forgot about. And when I built a Tetris clone as a test, Claude's version was "gorgeous with scores, next-piece preview, and great controls." ChatGPT made a basic clone. Gemini's worked but wasn't as polished.
Gemini 2.5 Pro is interesting though—it's not quite as good as Claude, but it's significantly cheaper via API. For high-volume coding tasks, that price difference adds up.
ChatGPT's coding has improved a lot with GPT-5, but it still tends to be more verbose and occasionally suggests deprecated approaches.
My pick: Claude for important code, Gemini for quick scripts and prototypes.
Writing: Claude, but ChatGPT for casual stuff
"Claude is the tool that best follows my instructions," and I 100% agree with this. Hand it a detailed brief and it executes. It matches tone, it hits word counts, it remembers constraints you mentioned three paragraphs ago.
For casual writing—emails, quick social posts, brainstorming—ChatGPT actually feels more natural. It's less formal, picks up on casual tone faster, and doesn't overthink simple requests.
Gemini's writing is... fine? It's accurate but tends toward the generic. I rarely use it for creative work.
My pick: Claude for anything important, ChatGPT for casual communication.
Research and current information
Gemini wins this category easily, and it's not close.
Gemini has real-time search built in. Its information is current. It can watch YouTube videos and summarize them. It handles massive documents without breaking a sweat (that 1 million token context window is real).
ChatGPT's browsing is hit-or-miss. Sometimes it searches, sometimes it doesn't, and you're never quite sure if you're getting current info.
Claude only recently got web search and it's... okay. Functional but not its strength.
My pick: Gemini for anything requiring current information or research.
Image generation
ChatGPT's DALL-E integration is solid. Gemini's Imagen is also good. Claude doesn't do image generation natively.
But honestly? For serious image work, you want dedicated tools like Midjourney or FLUX, not chatbot integrations.
My pick: Neither, really. Use dedicated image tools.
Multimodal (audio, video, vision)
Gemini 3 is the clear winner here. It can analyze audio files, process videos, and its vision capabilities are excellent.
ChatGPT's Advanced Voice Mode is impressive for real-time conversation, but Gemini handles more formats more reliably.
Claude's vision is good for images and PDFs but doesn't do audio or video.
My pick: Gemini for multimedia analysis.
Computer use and agents
Claude's "Cowork" feature is genuinely ahead of the competition here. It can control your browser, manage files, and actually execute multi-step tasks on your computer.
ChatGPT and Gemini both have agentic features coming, but Claude's actually work today.
My pick: Claude for anything requiring computer control.
Ethics and refusals
I know some people hate this category, but it matters.
Claude is the most restrictive. It'll refuse edge cases that ChatGPT and Gemini handle fine. This is annoying for creative writing involving conflict or mature themes.
ChatGPT has loosened up significantly this year. Most reasonable requests go through now.
Gemini is in the middle—fairly permissive but inconsistent.
My pick: Depends entirely on what you're doing. Claude's restrictions are a feature for some users, a bug for others.
The real answer
Stop looking for one "best" AI. The question isn't "ChatGPT or Claude?" It's "what am I trying to do right now?"
Here's my actual daily usage:
- Morning emails and quick tasks: ChatGPT (faster, more casual)
- Deep work and writing: Claude (more thoughtful, better execution)
- Research and current events: Gemini (real-time info, massive context)
- Coding projects: Claude first, GPT-5 for second opinion
- Analyzing documents: Gemini (handles large files better)
I know that sounds complicated. Managing three AI subscriptions IS complicated. It's also expensive—about $60/month across all three.
That's exactly why tools like LazySusan exist. One subscription, all the models, one interface. You pick the right tool for each task without managing multiple accounts.
Final thoughts
The era of "one AI for everything" is ending. The best approach is understanding what each model does well and switching based on the task.
If you can only pick one:
But if you can, use all of them. They're genuinely different tools now.
What's your setup? Do you have a clear favorite or do you switch around like me? Curious to hear how other people are handling the multi-AI world.