Model selection / task fit

We're all using AI for work now, and if you're like me, you maintain a subscription to ChatGPT, Gemini, and Claude. And that's because each model is the best at different things.

Well, we recently got a new Claude and a new ChatGPT model, and so I've been doing some more testing to figure out which one is best at each task that we do on a regular basis, and I just want to share what I found so far. Claude is still our go to model.

It's the one we use by default for most of our work. By far the best at writing, and I think the best at following instructions.

It also has the best harness, so it's a platform that we find most of our work living inside of for all of our projects and skills are set up. However, ChatGPT's codex is definitely catching up quickly.

Claude has been amazingly good inside of Excel and PowerPoint, but now ChatGPT integrates directly with Google Sheets. It's kind of sad that I'm so excited for ChatGPT to be inside of Google Sheets when Gemini is already there, but unfortunately, Gemini is just not that great at spreadsheets yet...

nswers to simple questions. If I'm looking for a recommendation for a new piece of software, I'm gonna trust Gemini the most at this point.

I should probably clarify I'm talking about the Pro version of each of these models. Otherwise, the only time I find myself using Gemini is when I don't want to spend the time or the tokens.

It's definitely the most efficient model right now. There are a lot of areas where ChatGPT and Claude are neck and neck right now, and I think it comes down more to preference.

For example, in coding or complex reasoning, or in handling large amounts of messy data. And then of course, ChatGPT is far ahead on image generation right now, but anything could happen in the next six months.

I honestly think all three of these model makers are well positioned to take the lead in any of these areas.