ChatGPT 4 Vision Vs Claude 3 Opus

I just asked ChatGPT 4 Vision to translate this website - https://www.ad.nl/tech/techbedrijf-vervangt-binnen-7-minuten...

It returned the dreaded 'Sorry I am unable...' message. Seconds later I tried the same in Claude 3 Opus and it gave me the English text in 5 seconds. What is going on with OpenAI? Weird.

How was the translation?

Yeah, that's the point. It was spot on.

OpenAI probably wants to do a big release, ChatGPT 4 vision sucks. They know that whatever they release, other competitors will rush to copy it, so they need to release something that has an excellent advantage to reap the benefits.

We have to wait and see when they release Sora and new stuff.

Claude is better at coding. I use these LLMs all day long, am very surprised to see OpenAI behind.

I already cancelled my OpenAI subscription until they come out with something new. They have absolutely no moat.

Any prompting tricks you've found to work well with Opus in particular?