A question I get asked quite a bit these days is “Which model is best?”
By model, I’m referring to a large language model, and it’s not the easiest question to answer because I think different models are good for different things.
But you can compare models performing individual tasks using a tool called Airtrain. It's super helpful for seeing how various large language models perform side-by-side.
This super quick video covers:
Setting up a free AirTrain account 💻
Choosing models to compare (I used Claude, Gemini, ChatGPT, and Llama)
Adjusting settings like model versions and temperature
Running a prompt across all models simultaneously
Analysing the results for quality and creativity
Exploring Generative AI Models
In this use case, you’ll see that Gemini was the winner of this challenge, but I have to say that the latest release of Claude (3.5 Sonnet) is really impressive, and I’ve been using it a lot lately. OpenAI (ChatGPT) still shines when using custom GPTs, and Perplexity AI is my go-to for research questions needing sources (e.g., property data like days on market).
Remember, there's no "best" model overall – it depends on your specific needs. That's why it's crucial to keep our options open and experiment with different models for various tasks.
Happy hunting! 🤖